In AIVaaS™, AI is approached as a business transformation and not as a technology deployment. The goal is an AI-Infused Business in which AI becomes part of how the organisation creates, delivers and captures value.

This publication has run two series. The first was five stories about a room where people argue about AI, and about Ana, a customer the systems knew only through data. The second was nine weeks of articles building AIVaaS™ element by element, the approach for turning AI ambition into value.

Before the second series began, the promise was that nine weekly articles would stand for nine months. Ana has reached the end of them.

This article connects the two series and tests whether the connection holds. The stories were written first, and they left questions on the whiteboard without answering any of them. The methodology answers them, one element at a time, and almost every question the room asked turns out to have an element with its name on it.

The room has a whiteboard covered in arrows. Someone says it looks like the London Underground. Harrison Chase agrees, and adds that the map does not determine the destination. It only makes sure you do not miss your stop. [LINK Story 1]

AIVaaS™ is a map of that second kind: three pillars, fifteen elements, a foundation beneath them, and one formula that connects it all. The nine articles walked the map element by element. This article walks it from the ground up, and stops six times.

Each stop opens with what the method says and who is accountable for it, then returns to the room where the same question was already on the board. Ana closes each one. Organisations invest in AI, and Ana keeps score.

New here? Ana and the room are introduced in Story 0. She is the reason the methodology exists, and the methodology itself opens here.

1. Governance begins by authorising ambition

In AIVaaS™, AI governance decides how far AI may go in creating value. Its first act is the authorisation of ambition. Ambition comes in four levels. At L1, AI supports people inside their existing tasks. At L2, it augments end to end processes. At L3, it drives outcomes the business could not produce before. At L4, it runs through the whole business strategy.

The critical leap sits between L2 and L3, because below it AI improves what the organisation already does, and above it AI changes what the organisation can offer, how value is created, and what competitors have to answer. That leap cannot be handed to a technology function. It changes the organisation’s understanding of value, risk, investment and future advantage, which makes it a leadership decision before it is a technical one.

Accountability for it belongs to the CEO, the board and the leadership team, and accountability of this kind is not delegable. The highest governance body is the board itself, extended with the roles that make transformation real: the Chief Transformation Officer, the Head of the Strategic Project Office, the Head of Enterprise Architecture, the Lead Business Analyst.

Governance holds all four terms of the readiness formula together, because transformation fails when the parts move at different speeds. One of those roles carries a map of its own. The Business Architect holds the design of the whole organisation and joins the business view to the AI technology view. They stand between the CEO on one side and the CAIO and HR on the other, because all three views have to end up in one picture.

The room asked this first. Story 1 opens with Harrison’s whiteboard and names its own limit: an architecture of movement, with no architecture of purpose behind it. The room had a map of how the agents move, and none of what they move through.

The same story ends with an empty chair, and gives the reason plainly. There is no business architect in the room, and nobody invited one. The story describes them as the person who knows why the company exists, and whether the orchestration supports that why or quietly erodes it. That is a description of the work governance cannot do without, written before any of the methodology was published.

Story 3 adds the other half. Mara, a synthetic persona built from a thousand customer interviews, answers precisely, with arguments nobody can immediately refute, and nobody checks whether she is right. She speaks with the authority of data.

AIVaaS™ answers with three tests of a reviewer. One who does not understand the output cannot produce judgement. One without authority cannot enforce it. A review step that records approvals and never disagreement produces neither.

So the reader has two things to check. Who knows the design of the organisation as a whole, and whether that design connects business requirements with AI technology well enough for a use case to stand on it. And whether the people signing off AI solutions can weigh the value they are signing for.

Ana appears in that article as the final test. Governance that produces a safer organisation the customer cannot feel has done only part of the job.

2. Value has an address

Value is relational. Something is valuable to someone, in some context, and an AI opportunity that cannot name the stakeholder it serves has no measurable KPI to stand on. Shorter handling time is value for a customer. Lower cost per transaction is value for an owner. Less repetitive work is value for an employee.

Prioritisation then becomes a decision about distribution rather than a ranking of attractive ideas. The Value Creation Matrix sets out four quadrants: Q1 cost savings, the hygiene factor that funds the rest, Q2 revenue growth and enhanced customer value, Q3 the radical productivity leap on the inside, and Q4 new forms of value, where differentiation and disruption live. Ambition sets the reach of the portfolio, because higher levels keep doing everything the lower ones do.

The portfolio is also the last checkpoint where the selection can still be steered, and it is where Ambition Fusion is verified in practice. Every opportunity that survives has to belong to the business strategy, because AIVaaS™ knows no separate AI strategy to retreat to. A portfolio drifting away from the strategy has already given its answer.

This is the work of the first pillar, AI Business Value Definition. Accountability sits with leadership together with the heads of business functions, and it produces the Business-Ready term of the formula.

The room reached the same point from two directions. Kambhampati looks at Mara and calls her the centroid of a data cloud, then names what a centroid can never show: the edges, the outliers, the customer who stayed loyal for ten years and left one day without a word.

Averaging a thousand customers produces a recipient who lives nowhere, and every KPI derived from that recipient measures nothing in particular. Mara was accurate throughout. She had no address.

Lakhani takes the marker in the same room and writes two words, Exploitation and Exploration, and says most organisations want the second and pay for the first.

AIVaaS™ supplies the mechanism. An organisation commits to L3, the portfolio fills with Q1 savings because those calculate most easily, and each one is approved on its own merits. A year later the portfolio says L1. Nobody voted against the ambition. The budget did, one approval at a time. And the observation gains an address of its own: an empty Q4 in an environment that demands transformation means no new business models are in design.

The reader can run this on a single artefact. Look at the stakeholder profile of the opportunity list. It reveals the actual ambition, whatever the strategy slide says.

Ana never sees the portfolio, the quadrants or the shortlist. She sees what reaches her.

3. The person who has to work differently on Monday

Value distributed on paper moves nothing by itself. AIVaaS™ reads every opportunity through the stakeholders it touches, one at a time, through four questions: what they gain, what they pay or risk in time, control or the job itself, how they see it, and whether their support is needed for the opportunity to work at all.

The third question carries the most weight and is skipped most often, because perception is what people act on. Underneath it sits trust, and trust is built from memory. Where earlier initiatives promised much and returned little to the people who delivered them, the next promise lands on that record.

The human veto is the quiet verdict that follows. The value of this AI belongs to the company, the saving belongs to the owner, the risk belongs to me. From where the employee stands, that verdict is rational, and it is exercised by withholding work rather than by objecting to it. Readiness at this stop is a state of willingness.

This is the work of the second pillar, Organisation Enablement. Accountability sits with HR together with managers and employees across the organisation, and it produces the Organization-Ready term.

And here the room says something about itself. Across four stories it held four external thinkers, a strategist in the corner, a synthetic persona, and the customer, first through a screen and then through a phone. It never once held the person whose work the orchestration changes.

The stories come close to naming it. Kambhampati, arguing for an AI critic on the team, says that when the critic is a machine nobody takes it personally, nobody is offended, and nobody goes quiet for the rest of the week after the meeting. Everyone in the room then thinks about the meetings they survived, where the truth was present and nobody said it out loud.

That silence and the withheld work are the same event at two sizes, and this is the lesson the stories were built to deliver. The veto is the decisive moment in AI transformation because it is exercised by the only person nobody asked.

A room can be full of people who understand how AI works and still be missing the one person who has to do the job differently on Monday morning. That room produces a plan that is correct on paper and thin in practice.

Lifting the veto runs on three moves held together: a vision leadership visibly owns, change management run through value rather than through steps, and an employee experience that raises engagement, because employees now see value that is theirs. AI literacy belongs here too, understood as judgement rather than tool training, since the human in the loop has to be able to think.

For the reader this is one action, taken earlier than feels comfortable. Before the next deployment is designed, find the person whose day it changes and ask how they see it. The answer changes the design, which is why it has to arrive while the design is still open.

Ana meets none of this. She meets a person, or a service a person built, and she feels whether someone cared. Employee experience and customer experience are two sides of one coin, and people who have quietly vetoed the change deliver her a thinner version of everything planned for her.

4. Readiness needs a why

Process & Data Readiness is business logic before it is a technical audit. Processes show where value should flow, where decisions belong and where customers are met. Data shows whether value was actually created and where needs surface, including the ones nobody has voiced.

Semantics gives the records their shared business meaning, and systems act on the meaning they are given. Where none exists, they act on probabilities.

The design of an AI use case then stands on two grounds. Its business ground is the Stakeholder AI Opportunity, which states what success means. Its architectural ground is the design of the organisation. Business Solution Design is the last point at which business intent can still govern the technical solution. [LINK A8]

This is the work of the third pillar, AI Design & Implementation, and accountability inside it divides in two. AI-Ready sits with the CAIO and the heads of business functions. Implementation-Ready sits with the AI architect together with the PMO, project managers, Business Analysts and implementation teams.

The Business Analyst carries the individual use case, from the need through the requirements into the design. The Business Architect carries the coherence check across the whole, which is the level the foundation works at.

The room supplied the evidence early. The first sentence of this publication describes a company that held all of Ana’s data and did not know her. Every record was accurate: eight years, loyalty score 94, a pharmacy transaction four days old, then three days of silence. The meaning running across those records was in no table.

Someone in the corner said two words. Call her. The agent listened, and Ana said the sentence that reorganised everything downstream: she will probably start buying completely different things now, and she does not know which yet.

Story 1 had already described the alternative. Andrew Ng is asked what happens if the instructions are wrong, and answers that the agent will then do the wrong thing exceptionally well.

A budget for data cleaning does not buy meaning. Map where the decisions live, and keep people at the ones where the unexpected arrives.

Ana does not experience an operating model. She experiences whether the business models made room for her new life.

5. The sum and the destination

Value matures through three states. It begins as potential value, the promise inside an opportunity. It becomes emerging value when design, prototypes and early use start producing evidence. It becomes realized value when a result arrives and lasts in daily business.

Each transition needs evidence, and evidence needs a baseline, criteria set in advance, and an honest view of how much of the result the AI solution may claim.

AI Solution Evaluation asks whether one solution delivers its expected contribution. Value & ROI Realization asks whether those contributions together move the AI ambition, and through it the business strategy.

The AIVaaS™ commitment is value for all stakeholders from the start, even when realization unfolds at different times. Where the commitment excludes some stakeholders from the start, the resistance returns later as the human veto.

A solution can reach every local goal and add almost nothing to the transformation. It can stay isolated, duplicate capabilities that exist elsewhere, or improve an activity while the operating model stands still. A sum of successful AI use cases is therefore not necessarily an AI transformation.

The portfolio has to be read as a whole, including enabling value: the reusable data, shared components and capabilities that later work will build on, which a narrow review would cut. Accountability sits with the CEO, the board and the leadership team.

Story 1 closed on this in one line, before any of the method was published. The risk is not that the orchestra plays wrong notes. It is that the orchestra plays the wrong piece. Every musician can be correct while the evening is lost.

In the article that covered this element, Ana was in month eight. The room prepared, the bag by the door, the route tested, everyone around her knowing who does what. She prepares because of the value that is coming, and every item on the list exists only because of it.

6. The one capability that must not become a routine

Readiness in AIVaaS™ is read at two scales, and the difference matters here. Business AI-Ready covers the organisation as a whole through the formula, with its four terms and their keepers. AI Solution Evaluation & Readiness covers one solution and the context it is deployed into. Capability sits below both, because it outlives every deployment.

Capability is the repeatable organisational ability to create and realise value across solutions, and most of it is business capability rather than technical. Continuous Capability Growth is the discipline that builds it, keeps it and extends it.

Reading the outside economy continuously, which the first article of the series already called an organisational capability and not a report. Fusing AI ambition into the business strategy and keeping the two aligned as both move. Finding the AI opportunities worth pursuing and choosing correctly among them. Designing new business models beside the running operation, which becomes a permanent capability rather than a one off exercise. Producing evidence that value arrived.

Trained people, a centre of excellence, purchased technology and a maturity score may support capability or describe it, and none of them proves it exists. It becomes real through routines, decision rights, institutional memory and the ability to transfer learning from one solution to the next. One question tests it. Would this capability survive if its main carrier left the organisation today?

It has to keep growing for two reasons. The forms and the scale of value the organisation pursues expand, and AI itself keeps moving. Accountability sits with leadership, with the Business Architect holding coherence across the whole.

And this is where the map runs out. In Story 4, after the call, Lakhani draws a circle and writes Presence inside it. Then the marker changes hands and a fourth circle appears: the capability to be surprised. The reason follows immediately. Surprise cannot be made into an algorithm, because the moment it is, it becomes expectation.

Every other discipline in AIVaaS™ exists to make good results repeatable, and to repeat them each time at a higher level. That rise is the fourth wall of this room, the spiral: the same questions returning from a higher level of understanding, with more to build on.

This one has to be protected from repetition. It is carried by a decision, taken by a living person, every single time, at the point where the system says that it did not know this and the organisation treats that as a beginning.

There is no element on the map that can take it over. That is the boundary the method draws around itself.

In the article that covered this element, Ana was days away. The value she had carried for nine months was about to get a face, and nine months of growing capabilities were only the beginning, because the child will keep asking for abilities she does not have yet. Organisations are no different.

Ana gave birth this week

Which is where her sentence on the phone comes due. A life reorganising itself into new needs, before any product has a name, and now the needs multiply and no single organisation covers them alone. The paediatrician holds one part of the picture, the pharmacy another, the nursery a third.

That is where this publication goes next. First to the process levels, and the question of how deep in the process hierarchy an agent system has to be designed before it understands what a customer treats as value. Then to the level above a single organisation, where value is created for the customer simultaneously by several actors, and to the canvas built for designing it.

Then back to AIVaaS™ itself, to the readiness formula, and to the roles this article has named without unpacking: the Business Architect, the Business Analyst, the Process Analyst.

Before that, a pause.

The child in this series is not mine. The methodology is. Nine weeks of articles were its nine months, and this week both arrived. I am taking a month of paternity leave.

In AIVaaS™, AI is approached as a business transformation, not merely as a technology deployment. The goal is an AI-Infused Business in which AI becomes part of how the organization creates, delivers and captures value.

Deployment feels like the end of the road. The solution is accepted, signed and running in real work. The teams that carried it move to the next one, and the organization counts another AI success.

Two decisions refuse to close with the deployment. The operating solution stays under evaluation, and the AIVaaS™ foundation keeps asking what no deployment can answer.

Was the Stakeholder AI Opportunity realized? Did its value advance the AI ambition, and through it the business strategy?

No pillar can answer that alone. The first defines the value to pursue. The second prepares the organization to pursue it. The third turns the pursuit into working solutions. Confirming what value truly arrived is a different kind of work.

Below the pillars lies the ground they stand on: AI Governance & Value & ROI Realization & Continuous Capability Growth. Every article so far has drawn on it through the element it covered, and governance has had its own chapter. As a whole, the foundation has stayed in the background of the story.

Whatever the pillars build, the proof of its value lives below them.

Three questions that never close

The pillars work as one system. Most of their elements connect, complement and build on one another, and the work returns to earlier elements whenever the transformation demands it. Each element has a moment when its role is complete.

The foundation’s role is fulfilled only by never stopping.

It runs on three questions. Is the promised value truly arriving, and is it moving the ambition and the strategy? Is the organization growing the capabilities to create ever more value, and to keep pace with AI itself? And who holds the measures and the boundaries while the answers keep changing?

The third question belongs to AI Governance, the discipline that decides how far AI may go and holds leadership accountable for the answer. The first two belong to Value & ROI Realization and Continuous Capability Growth, and they begin before any AI solution is designed.

The promise comes first

Value & ROI Realization works through the whole transformation. It asks how value is defined, targeted, evidenced and governed before, during and after AI implementation.

Its first act comes at the moment a Stakeholder AI Opportunity is born.

Every Stakeholder AI Opportunity carries two connected promises: value for named stakeholders and a defined contribution to the AI ambition. Each opportunity is one piece through which that ambition is expected to become real, and it is judged worth the investment on both promises.

That evaluation of expected value happens at the start of the road, and the discipline then follows both promises all the way to the result.

The AIVaaS™ commitment is value for all stakeholders from the start, even when realization unfolds at different times.

If the commitment excludes some stakeholders from the start, resistance can later become the human veto on the transformation.

From promise to result, value matures through three states. It starts as potential value, the promise inside the opportunity. It grows into emerging value when design, prototypes and early use begin to provide evidence that the promise can be fulfilled. It becomes realized value, a result that has arrived and lasts in daily business.

Each transition needs evidence, and evidence needs preparation: a baseline before the change, criteria for the expected value, and an honest view of how much of the result the AI solution may claim.

Technical evidence shows how the solution behaves. ROI can be assessed only when that evidence is connected to business and financial outcomes, across the time lag between deployment and effect.

From contribution to outcome

AI Solution Evaluation determines whether an individual solution is delivering its expected contribution to value. Value & ROI Realization determines whether those contributions are sufficient to realize AI ambition and move the business towards its intended transformation.

In fragmented AI adoption, the gap between the two can stay hidden for a long time. A solution can reach its local goals and still add little to the transformation.

It can stay isolated. It can duplicate capabilities that exist elsewhere. It can improve an activity while the operating model stands still. It can create local savings that serve no chosen ambition, and leave the business model and the strategic position untouched.

That is why a sum of successful AI use cases is not necessarily an AI transformation.

Technical performance is evidence. Solution value is a contribution. Transformation value is the destination.

The AIVaaS™ approach changes this pattern. Every selected Stakeholder AI Opportunity is expected to carry its share of the transformation, local results included.

The Value Creation Matrix shows which form of value each opportunity is expected to create and for whom. What remains open is how the shares interact and what they add up to against the planned value.

More than a sum of returns

Value Realization therefore reads the whole portfolio. Do the contributions reinforce one another, or perhaps even work against each other? Does the portfolio cover the chosen level of AI ambition? Which AI investments deserve to continue, grow, connect or stop?

The portfolio view changes single decisions too. A solution may return less direct ROI than expected while building reusable data, shared AI components and capabilities that later work builds on. This enabling value belongs in the judgement, or the review cuts what the next opportunities need.

Value is also read across stakeholders, because every form of value has an address. Each Stakeholder AI Opportunity is judged with its stakeholders first.

Was their need truly covered? Did the promised value appear, and was it delivered and sustained? Is it compatible with value created for other stakeholders, or does it erode it? And does the combined effect support the business ambition?

All of it flows back to where the road began. Value Realization confirms each opportunity’s contribution to the AI ambition, and through the ambition its contribution to the business strategy. What it learns feeds the next opportunities, the next investments and, when needed, the ambition itself.

The Critical Leap: Where value realization changes its logic

Before value can be realized, leadership must choose what kind of value the organization intends to pursue.

The choices line up on a spectrum, from cost reduction on the left to net positive impact on the right (the Value Spectrum). The spectrum shows a widening scope of value: who receives it, how much the business must change to create it, and how much capability integration is required to sustain it.

For many organizations, the left side is the natural starting point. Cost reduction, increased profitability and revenue growth can often be reached within AI adoption, sometimes through strong individual AI use cases and sometimes through a disciplined portfolio of them.

The Critical Leap marks the point where that logic changes. To the right of the leap, sustained value realization at scale depends on connected capabilities across solutions, data, processes, decision rights and ways of working.

It aligns with the boundary the AI Ambition Matrix draws between L2 and L3: the move from AI adoption to AI transformation.

What changes is what the organization must be able to do.

Organizations enter the spectrum at different points and may choose to leap. The further right they aim, the more their value realization depends on capabilities that reinforce one another, scale and continue to grow.

Choosing a form of value is therefore a leadership decision before it is a technical one. The chosen form of value, the level of AI ambition and the required capabilities must stay aligned.

An organization cannot remain at L2, build only for adoption and expect right-side value to materialize.

What remains and keeps growing

What the organization must be able to do beyond any single solution has a name in the AIVaaS™ foundation: capability. Continuous Capability Growth is the discipline that builds, preserves and extends it.

Readiness is the state of a specific solution and its environment. Capability is the repeatable organizational ability to create and realize value across solutions.

Readiness is assessed for a specific solution and its deployment context, inside the third pillar. Capability outlives each deployment, which is why it belongs to the foundation.

An organization can be ready for one solution and still be unable to repeat the achievement. Capability makes value creation and realization repeatable, and it must keep growing for two reasons: the forms and scale of value pursued expand, and AI itself keeps evolving.

Capability is easily confused with the assets that support it. Trained people, a centre of excellence, purchased technology and one-off training may support capability. A maturity score may describe it. None of them proves that the capability exists.

Capability becomes real through routines, decision rights, institutional memory and the ability to transfer learning across solutions.

Would this capability survive if its main carrier left the organization today?

Organizations often stumble later, when attention shifts to the next problem before learning from the previous solution has been absorbed and reused.

Every loop left open creates a double loss: unrealized or unconfirmed value, and the capability the solution should have left behind.

That is why capabilities are built alongside solutions and, where the ambition demands it, before them. The capabilities built around a solution are part of its enabling value, and part of what the portfolio view must protect.

Capability is what the organization retains from one solution cycle and can reuse in the next. It makes the next leap faster, more reliable and less costly.

Left of the Critical Leap, capability is an advantage. Right of it, capability is the ground on which sustained value stands.

Who holds the measures

One discipline holds the other two together. AI Governance sets the measures, boundaries, decision rights and accountability within which Value & ROI Realization tests outcomes and Continuous Capability Growth renews the organization’s ability to create them.

Those boundaries travel down to every solution. Business Solution Design translates them into the solution. Where needed, the AI Solution Constitution codifies the non-negotiables, and the harness makes them executable.

Three disciplines, one foundation. Value & ROI Realization confirms what value arrived. Continuous Capability Growth keeps the organization able to create more. AI Governance holds the frame in which both continue.

The pillars build solutions. The foundation keeps value provable, capability growing and decisions governed.

Ana and the life about to begin

Ana is days away. Her life is about to change in a way she could prepare for and cannot yet measure.

The value she has carried for nine months is about to get a face. Everything she has ever counted as valuable will rearrange itself around it, and the meaning of this value will keep growing with the child.

Nine months of growing capabilities were only the beginning. As the child grows, it will keep calling for abilities Ana does not yet have.

Organizations are no different. The value they create will keep asking them to become capable of more. And the customers Ana represents will keep expecting that value to grow with their lives.

In AIVaaS™, AI is approached as a business transformation, not as a technology deployment only. The goal is an AI-Infused Business in which AI as technology becomes part of how the organization creates, delivers and captures value.

Every AI initiative lives in two worlds. The business world speaks in stakeholder needs, decisions and value. The technical world speaks in models, data, proofs of concept and traces. Between them lies a distance, and most of the value that AI promises is lost somewhere along it.

The two worlds often share the same words: validated, evaluated, ready, accepted. Each word carries a different meaning on each side, and behind each meaning stands different evidence, a different decision and a different accountability.

Behind the words sits a clean division of accountability. The business side answers for value created for the customer, the core of every business model. The technical side answers for an AI solution that works.

Between the first accountability and the second stretches a long distance, and it is a competitive one. The company that crosses it faster holds a shorter distance to the customer than a competitor fighting for the same one.

From intention to realized value, a chain of questions waits along the road. Each question has a business face and a technical face, and each becomes dangerous when only one side answers it. Answered together, the two faces form one decision logic, and that is what shortens the distance.

Start right, or the distance grows

The road to value begins at a Stakeholder AI Opportunity: a business object and a value goal, a partial manifestation of the AI ambition, with its own place among the types of value the organization pursues (its position in the Value Creation Matrix).

An opportunity answers one question: where, for whom and through which business change could AI create meaningful value? It arrives with owners, stakeholders and an expected value attached, long before anyone mentions technology.

This is where evaluation begins. The first object under judgement is the opportunity itself: how important it is, which stakeholders it serves, what value it should create, how well it fits the strategy, and whether it deserves further investment at all.

An AI use case is a designed path toward the opportunity: a concrete combination of people, processes, decisions, data, tools and AI. It is born in design, and this order matters. The opportunity states what success means; the use case shows one way of reaching it.

One opportunity may need several use cases before its value is within reach, and no single use case defines the whole opportunity. Losing sight of this relationship is how value quietly narrows: the use case gets delivered, the opportunity stays unrealized.

Most business conversations about AI start with a list of use cases to identify and prioritize. Read closely, most entries on those lists are opportunities in AIVaaS™ terms: places where AI might create value, named before any path exists. One word carries both objects, and the goal disappears inside the path.

The business model behind the opportunity has already promised value: created for the customer, delivered through the offer, captured in the results. The use case takes that promise to the test. And when the transformation is later measured, the question will return to the beginning: was the opportunity realized?

Where the two worlds must meet

Design is where the two worlds first sit at the same table, because the object of design belongs to both. What travels toward deployment is a business solution: people, processes, decisions, data, tools and AI working as one system, with the model inside it.

The business half of design has one central duty: to keep the expected value of the opportunity alive inside the solution. Every trade-off made at this table is measured against that value, or it starts working against it.

Business analysis builds the bridge and keeps the journey fast and on course: every detour off the road makes the distance longer. It clarifies the need, then defines the business, stakeholder, solution and transition requirements that carry stakeholder needs and expected value into design.

Design answers with a possible shape, and the shape reveals new requirements, so the loop runs until both worlds recognize the solution. The table fills with trade-offs along the way: value against cost, usability against feasibility, AI autonomy against human control, provider standardization against the actual business process.

The name itself can mislead. Solution Design sounds like a technical discipline, which is why AIVaaS™ insists on Business Solution Design: a solution here always contains two poles, the business system and its technical implementation.

The word business in that name works on two levels. The constitutional level is inherited from the foundation: how far AI may go, what it must mean, what the transformation owes its stakeholders. The solution level is written here: the value hypothesis, criteria, decisions and business logic of this one opportunity.

Business strategy is the constitution of a company. An AI Solution Constitution, an AIVaaS™ term, plays the same role for a single AI solution: non-negotiable purpose, principles, priorities, business rules, decision boundaries, autonomy limits and escalation paths, written down before any technology enforces them.

The constitution governs. The agent harness makes it executable in daily operation. And the evaluation harness, the machinery that runs repeatable tests and records how the solution behaves, will later produce the evidence.

The last line of defence

Business Solution Design is the last point at which business intent can still govern the technical solution. After it, implementation begins, and every trade-off becomes more expensive to reverse. What the line defends is the value the opportunity promised.

The pressure at this line is real. Providers naturally lean toward their own solution patterns, clients want progress, and attention slides toward the solution itself. The stakeholder need gets overlooked one small decision at a time, while everyone in the room believes they are still serving it.

A design built from one side carries a double price: higher cost and lower realized value. The road then gets walked more than once, redesign after redesign, or it reaches its end without the main thing: satisfied customers willing to take over the result. Those customers can be external, internal, or both at once.

The numbers behind this are brutal. IDC found that 88 percent of AI proof-of-concepts never reach production, and MIT research put the share of AI pilots failing to deliver expected returns at 95 percent. Behind many of those failures sits the same historic mistake: the client handed the whole solution to the provider.

Ownership of the business problem, the process design, the division of work between humans and AI, the success criteria and the value judgement stays with the business. Everything else is collaboration: the end customer, the client and the AI solution provider weaving their separate accountabilities into one result.

One more thing is written at this table, and its timing decides everything later: evaluation criteria, acceptance criteria and test scenarios are designed with the solution, the same way good test scenarios are written at design time, long before testing. Transition requirements are identified here too, and readiness will come asking for them.

The third element of the pillar now has its full name: Designing, Prototyping & Validating AI Use Cases. Three actions share one object. Design gives each use case its form; the next two make sure the form deserves to move on.

A prototype makes the designed use case visible before change becomes expensive: assumptions, interactions, roles, decisions and expected behaviour stop living in documents and start acting in front of people. What it tests first is the value hypothesis, with technical feasibility riding along.

The prototype is the method. Validation is the judgement it supports: are we shaping the right AI use case and the right Business Solution for the right stakeholder need and intended value? We evaluate the prototype in order to validate the use case, the design and the value hypothesis behind them.

This is why the word Validating stands in the scheme itself: the right solution gets confirmed here, where a wrong turn costs a redesign instead of a failed deployment. What leaves this element is a validated use case, shaped into a business solution and ready to face harder evidence.

From evidence to signature

By now the goal is in sight. The use case is validated, the agent harness has turned the constitution and the design into something that runs, and the end of the road feels close. One thing still has to be confirmed: whether the value for the customer travelled the whole way with us.

The pillar meets the solution with its fourth element, AI Solution Evaluation & Readiness, and one question in four layers: can this solution be trusted with real work?

Verification comes first: was the agreed design built? On the technical face, implementation is checked against architecture, data flows and controls. On the business face, something subtler: whether the agreed roles, rules, decision rights and responsibilities survived implementation.

Evaluation then produces the harder evidence. The evaluation harness runs repeatable tests and records traces, and the business reads that evidence against one measure: the expected contribution to value, the limitations inside the solution and around it, and what would increase the value it delivers.

Readiness widens the view from the solution to its environment. The transition requirements written at design time return here as a checklist of reality: changed processes, prepared people, clear ownership, data in place, monitoring and support ready to receive the solution.

Readiness is a state. Acceptance is an accountable act. A named business owner authorises deployment, accepts the residual limitations and risks in writing, and can still say No. The human veto over AI transformation gets institutionalised here, solution by solution.

Only after the signature does AI Solution Deployment move the solution into real work, as the closing step of the pillar. The order is the message: evidence before readiness, readiness before signature, signature before deployment.

Every decision has two faces

The road can now be read as a chain of decisions. Each row below is one of them: the business meaning that must hold, and the technical evidence that must support it. When a meeting uses one of these words, this is the pair it should mean.

The first seven decisions live inside the pillar and close with it. The last two run beyond it: the operating solution stays under evaluation, and the foundation keeps asking whether realized contributions advance the ambition and the strategy.

Evaluation moves in with the solution

Deployment closes the pillar and changes the context of evaluation. Before it, evidence served the Go decision. After it, the operating solution produces evidence no prototype could: real drift, real incidents, real cost, real behaviour with real customers. Evaluation simply moves in with the solution and keeps working.

The first thing under watch is the transition itself: did the changes around the solution really happen, above all the ones that depend on people? Do the new processes run smoothly? Do customers feel only the positive side of the change?

When those answers weaken, the solution walks back to a new readiness decision. Realized value is the only proof that counts.

Ana is almost there

For Ana, transition requirements are something different. Month eight is what they look like in a life: the room is prepared, the bag waits by the door, the route is tested, and everyone around her knows who does what when it begins.

None of it is the point. She prepares because of the value that is coming, and every item on her list exists only in its name. The day itself will choose its own time; her work is that everything is ready before it.

Organizations rarely stop to ask how much of a realized AI opportunity reaches the customer as value. They should. Ana will be watching more closely than ever, and where her attention ends, her AI agent’s begins.

Ana may forget. Her AI agent will not, and next time it will simply offer her a different choice: someone with a better understanding of her and her new needs.

In AIVaaS™, Designing New Business Models is not a creative exercise reserved for startups. It is the element that decides whether AI changes only how the business runs, or also the value it creates, and it begins far earlier than its place in the third pillar suggests.

The previous article ended with a business that was ready. Processes mapped to their decision points, data that caught the signal, a human at the decision that mattered. All of it prepared so that customers will be served well.

But does that last part hold? The processes and the data the previous article made ready belong to one place: the operating model, the way the business runs from day to day and what it offers as it runs.

Whether any of that movement also changed the value the business creates, and for whom, is a question readiness cannot answer.

The boardroom question is not only whether AI will make the business more efficient. It is whether AI will change what the business is worth, and to whom.

This article opens the second element of the third pillar, Designing New Business Models. The element sits late in the methodology. That placement hides how early its work begins.

How early depends on AI ambition. Organisations aiming at L3 (AI-Driven) or L4 (AI-ALL IN), where AI drives outcomes the business could not produce before, meet business models at the strategy table, inside Ambition Fusion.

Organisations that entered at L1 (AI-Support) or L2 (AI Add-On) meet business models here, at the latest. For many boards this element is the final checkpoint where anyone asks whether the change in operations has quietly moved the business model too.

Even below the leap the question has weight. A manufacturer takes AI through sales and production and leaves logistics untouched. The gains are real, and unrealised: its customer measures value in deliveries that arrive in time for its own customers, at the address the programme never reached.

The organisations pulling ahead read this early. They run today’s operating model and build tomorrow’s business models beside it, in parallel, and the distance they gain compounds.

Business models were planted in this series long before this article, in its first economic reading, and every station since has returned to them.

Hidden in plain sight

The opening article placed Designing New Business Models on the map, as the second element of the third pillar, with one line attached: at higher ambition levels AI enables new ways of creating value.

The first economic reading went further. In the comparison of three economies, business models stand among the eight dimensions the shift moves: they change how value is captured.

The four levels of AI ambition lean on them harder. What separates L1 and L2 from L3 and L4 is whether business models move. At L4, the whole rests on a portfolio of business models that would not exist without AI.

The Value Creation Matrix runs on the same axis. Its lower row holds value created on the operating model, its upper row value created on the business model, and the critical leap runs between the rows.

Governance met the term from above. The deeper governance act is how far AI may go: support existing work, improve the operating model, reshape the value proposition, or enter the business strategy itself. Two rungs of that ladder live inside the business model.

And the previous article closed at its door: customer needs read from data, including needs no customer has voiced yet, are where new offers and new business models begin.

Six articles, and the term carried weight in every one, always in a supporting role. It stayed there because business models belong to the vocabulary every executive is assumed to know. That assumption deserves a test.

What a business model is

A business model is the rationale of how an organisation creates, delivers and captures value, for its customers and for itself. Three verbs, one noun, and the noun does the work.

Most management teams carry the term in the singular: the business model, one per company, decided long ago. The singular is the first thing to correct. Every organisation runs several business models at once, one for each customer segment and each distinct part of the offer.

The count settles a quieter fear. A new business model is added next to the existing ones, and the existing ones keep earning while it grows. Designing new business models threatens nothing that works today.

The definition and the working tool both come from Alexander Osterwalder. His Business Model Canvas describes each model in nine blocks and gives the several models of one organisation a shared page.

Five blocks face the market: customer segments, the value proposition at the centre, channels, customer relationships, revenue streams. Four face the engine room: key activities, key resources, key partnerships, cost structure.

The centre deserves its own instrument. The Value Proposition Canvas zooms into the fit between what the business offers and what the customer is trying to get done: the jobs, the pains, the gains. Where the two profiles meet, customer experience is won, and with it the value.

A second definition sharpens the horizon. Gennaro Cuofano reads a business model as a systematic way of unlocking long-term value while delivering it to customers and capturing part of it in return. The long term matters: models are designed to outlive the tools that enable them.

And the operating model? It has been on the canvas the whole time, in the engine room blocks: the key activities, key resources and key partnerships through which the value proposition gets delivered.

The two models divide the work.

The operating model holds how the business runs and delivers, its processes, people, systems and what it offers day to day. The business model reads the whole above it: which value, for whom, and how the value returns.

This is the test the opening question stands on. A change in operations enters the canvas through key activities and key resources. Whether it changed the business model depends on what moved above them: the proposition, the segments, the revenue logic.

Business models decide who disrupts

Osterwalder sorts innovation into three types. Efficiency innovation improves processes and the existing business model. Sustaining innovation extends value propositions into new channels and geographies. Transformative innovation builds new business models and new growth engines.

The sorting predicts the outcome. Disruptors work in the transformative class, where new models are born. The disrupted work in the efficiency class, polishing the model they already have, right up to the day it stops carrying them.

Behind the types sits an older pair. Exploit names the short horizon: efficiency, control, more profit from what exists. Explore names the long one: discovery, new revenue, new markets. L1 and L2 exploit, L3 and L4 explore, and the critical leap is the crossing between the two logics.

The pair is drawn with a plus sign, explore plus exploit, and the plus is the strategy: the strongest organisations refuse to choose between the two.

The third stage of digital business models

Digital business models divide into three stages. In the first, the organisation creates, delivers and captures value with the help of digital technologies: the model is enabled by digital, and the value itself can still be touched.

In the second stage the form of the value itself is digital. SaaS, Data as a Service, marketplace platforms, Direct-to-Fan models such as Substack: what the customer receives, uses and pays for exists as digital value, and the whole logic of the model follows it.

The third stage is opening now: digital business models based on AI. One of its clearest emerging forms is AGaaS, Agentic-as-a-Service.

Here digital value turns active. It delivers outcomes through work performed on its own, and the intelligence itself becomes the product: the next evolutionary stage of the digital business model.

For every company the question stands: which of its business models will AI enter, and which new ones will it make possible.

Where business models enter the AIVaaS™

Defined, the term returns to the methodology with weight. The first pillar touches business models before they become implementation work, and the third pillar inherits what the first has, or has not, prepared.

External Strategic AI Impact Analysis reads other people’s business models first: which competitors have moved to AI-native models, which adjacent industries can now enter, and where the value pool is migrating.

AI Ambition Alignment then chooses a response. Choosing L3 or L4 is choosing to design new business models; an ambition above the critical leap with no model design attached is a wish with a number on it.

AI Strategy Fusion keeps the design where it belongs. Business models are where strategy takes economic form: choices about value, customers, capabilities and revenue become a working model. AIVaaS™ knows no separate AI strategy; new models arise inside the AI-Infused Business Strategy.

Defining Stakeholder AI Opportunities gives each model its addresses, because opportunities in the two upper quadrants of the Value Creation Matrix are business model opportunities by definition. AI Investments Prioritization then distributes the portfolio between exploit and explore. An empty Q4 in an environment that demands transformation now has a second name: no new business model in design.

The Foundation validates the model before deployment. Value & ROI Validation runs from the first pillar on: created, delivered and captured value are named and tested while the model is still a design.

AI Governance gains a portfolio to govern: which models are added, which coexist, and which are retired. Adding a model changes what the business is, and that decision belongs to leadership. Continuous Capability Growth closes the loop, because business model innovation with AI is now a permanent capability, not a one-off exercise.

The leap runs through the business model

The critical leap between L2 and L3 has appeared in this series as a leap of ambition, of posture, of governance and of readiness. This article adds the shortest description it has: the leap is where the business model starts to move.

What moves with it is the competitive position. Below the leap the organisation plays defence: gains on the operating model protect the market position it already holds, and every competitor is running the same programmes.

Above the leap the position itself starts to move: towards differentiation and competitive advantage at L3, towards disruption and market leadership at L4. The reward of real AI transformation lives there, the breakthrough to the leading position in the market and with the customer.

The levels are levels of ambition, and no rule says they must be climbed in order. Because a new business model rises beside the running operation rather than through it, an organisation can hold its position at L1 and L2 while building at L3 or L4 in parallel.

For organisations aiming at L3 or L4, the design work belongs to the first pillar, inside the strategy and the validation that runs with it. For organisations at L1 and L2, this element is the last checkpoint: the place to ask, once, what the readiness work has moved.

The answer can be nothing, and nothing is a legitimate answer. The check exists because the other answer arrives uninvited: processes and data redesigned around new decisions can shift a value proposition without anyone deciding they should.

Either way, the question has left the technical floor. What the business is worth, and to whom, is being decided again. And the first to register the outcome is not anyone inside the organisation.

Ana feels it first

In the stories that opened this series, Ana is the customer the organisation serves today, and the child she is carrying is the customer of the next, more intelligent economy.

Ana does not wake up with a wish to buy something. She wakes up into a life that changed three days ago, with needs she is only beginning to sense. Value, for her, starts in the need.

Everything this article has described, the canvas, the stages, the portfolio, is invisible to Ana. What reaches her is the answer itself: whether a need of hers is met, how completely, and at what cost in money, time and trust.

The product she meets comes from the operating model; the value she experiences, from the business model. When only the operating model improves, a need she already had is answered a little better.

A new business model reaches further. It answers a need that had no answer, sometimes a need she has not yet voiced, and she registers that before any internal metric does.

That is what her sentence on the phone meant. Buying completely different things describes a life reorganising itself into new needs, before any product has a name. The businesses designing new models around those needs are the ones that will exist in her new life.

Ana does not experience an operating model. She experiences whether the business model has made room for her new life. That is why the critical leap runs through the business model, not around it.

In AIVaaS™, Process & Data Readiness is not the technical audit most writing describes. It is the business logic of processes and data that determines whether AI improves isolated workflows or transforms how the business understands customers, makes decisions, creates value and operates end to end.

Boards no longer ask whether prepared processes and data matter for AI. The conversation has moved on: what is missing, and when will we be ready?

The answers arrive in a technical register: data quality scores, integration roadmaps, process documentation, a programme plan with milestones and a budget line.

Then one voice asks: AI-ready for what?

The question returns to the board. Readiness only has meaning against the level of AI ambition the business has chosen and the strategic directions it has committed to. Without the why, the what, the how and the when have nothing to lean on.

Naming the why is leadership work, the same accountability the previous article placed with leadership for how far AI may go .

The why is one half of what must come before the execution layer. The other half is the connection between processes and data, which most readiness models read separately.

Processes show where value should flow, where decisions belong and where customers are met. Data shows whether value is actually created, which decisions are made or missed, and where customer needs surface, including the ones nobody has voiced yet.

Processes and data give value to each other. Read together as business logic, they make the why testable: the value the business has committed to has to show up in real processes and real data before anything turns into use cases, prototypes and build decisions.

Readiness has a reference point

Readiness is a state of preparation, and preparation is always for something. In AIVaaS™ that something has two names: the level of AI ambition, set through Ambition Fusion, and the strategic directions committed in the AI-Infused Business Strategy.

The reference point reaches wider than AI. A new business model, a new market, a sustainability commitment, an acquisition to integrate: every strategic direction eventually lands on the same two carriers, the processes the business runs and the data it holds.

At pilot scale the missing why stays hidden. The first two levels of AI ambition improve work that already exists: AI supports people at tasks or augments running processes. The present operation serves as the reference point, and technical readiness can carry the effort.

From the third level on, AI drives outcomes the business has never produced before. Their reference point does not exist in the present; leadership has to name it before the work of readiness begins. The wider the change AI brings, the earlier the naming has to happen.

When the why stays unnamed, the loss is quiet: months, budget and the goodwill of capable teams, spent on precision without direction. Operational teams feel it first. They carry the task of readiness, and the starting point they need sits with the board.

The 2026 evidence points one way. PwC’s AI Business Predictions put technology at roughly one fifth of an AI initiative’s value; the rest is created in redesigned work. Deloitte and BCG find the same pattern among the leaders: redesigned roles, decision rights and end-to-end processes.

Readiness in AIVaaS™ is more than this article’s element. The whole preparation is one formula: Business AI-Ready = Business-Ready + Organization-Ready + AI-Ready + Implementation-Ready.

The formula exists to protect the why. Each term checks readiness against the same reference point, so the ambition named at the start is still the ambition being built at the end. Process & Data Readiness opens its AI-Ready term.

An organisation has more process levels than it thinks

Ask for the process map and most organisations bring workflows: task sequences in an IT system, an approval chain, a wizard guiding a user through steps. Useful, real, and the smallest part of the picture.

Processes run on five levels. At the top sit the main business processes: the end-to-end value streams of the business (from order to cash, from idea to product). Below them, the value-adding phases where customer value takes shape, and under those, the individual processes a manager can own.

The fourth level, subprocesses, breaks a process into measurable steps, with rules, metrics and decision points. Only the fifth level, the tasks, holds the work people and systems execute in sequence. That is where workflows live.

A workflow executes, and every task in it has a performer responsible for it. What it lacks is management. Its goals, its rules and the monitoring that notices a wrong outcome and corrects course all live on the levels above.

Most AI adoption enters here, into tasks and workflows, and inherits their limits. It executes faster, drafts faster, routes faster, while the decisions above stay untouched. In ambition terms, that is L1, AI-Support. Workflow entry is enough for it.

L2, AI-Add-On, already asks for more. AI layered over an end-to-end process improves only what the business sees whole: the value stream, with every phase and handoff inside it.

Above the workflows sits the business logic of processes, and it faces two directions. Inward: ownership, business rules, decision points, and the management loop of plan, monitor, correct.

Outward it faces the customer journeys the processes exist to serve. A journey reads as a process too: stages and touchpoints, each backed by an end-to-end flow, each a moment of truth where value is won or lost.

Decision points deserve special attention. They sit before and after workflows, deciding what enters them and what their output triggers. A workflow executes a decision made elsewhere. Where that decision happens, and on what basis, is process knowledge AI will need.

One management axiom seals the point. W. Edwards Deming taught it; Russell Ackoff formulated it most sharply: parts made to perform as well as possible, each on its own, do not produce a whole performing as well as possible.

Workflows optimised one by one do not add up to an optimised value stream. And the value stream reaches beyond the organisation: the whole a customer experiences is the junction of the customer’s processes with ours. Readiness that stops at the company boundary prepares half of the whole.

Data that does not know what it means

Where the customer’s processes meet ours, data faces its real test: the same accurate records can carry two opposite decisions about one customer.

Ana, the customer this series has followed from its opening stories , is about to be on the receiving end of one of them.

Ana has been a customer for eight years. Then, for three days, silence. The organisation holds every transaction she has ever made with it, and one of the latest is a pharmacy purchase: a test.

An agent reads the silence as a reactivation case: high probability of return, queue a discount. Every value in its records is accurate. The meaning that runs across them, a life about to change, is not a field in any table.

The usual response is technical: better quality, cleaner pipelines, a unified platform. All of it matters, and later articles will give the technical side its space. None of it changes what the records mean.

In AIVaaS™, the business logic of data comes first, and it stands on three members: the decisions the data must serve, the customers it must reveal, and the semantics that makes both possible.

The first member connects straight back to the process map. The decision points identified there define which data matters, at what moment, in whose hands. In the scene above, the decision was whether to send the discount or to pick up the phone.

The decision member carries two blind spots, both business rather than technical. The first is what the business does not know: signals never collected, signals held and never read. No system knew what the test had shown. That knowledge arrives only when the business is present to receive it.

The second is sharper: the decisions that should have been made and were not. Systems record executed transactions. The call never placed, the option weighed and dropped, leaves no trace. Gartner’s 2026 work on decision traces and context is an early attempt to capture this.

The second member is the customer. Steve Jobs liked to say that people often do not know what they want until someone shows them. Data with meaning is how a business shows them: read against the journey, it surfaces needs forming before the customer voices them. Ana’s records held one.

The third member, semantics, makes the first two possible. It is the shared business meaning of what the data describes: what a customer is, what an order is, what a silence is, and how they connect across systems.

Gartner now ranks the semantic layer as critical infrastructure for AI. The reason fits one sentence: systems act on the meaning they are given, and where none exists, they act on probabilities. A discount for a life moment is what that looks like.

The question for the data organisation grows. Clean and complete stay necessary; the bar above them is meaning: which decisions the data serves, which customer needs it reveals, which value it makes possible. That is data readiness read from the why.

The layer in between

Business logic above, technical execution below. The two layers of preparation meet in the middle, and that middle is where organisations see their processes and data as one for the first time.

Process mining reads the digital traces work leaves in IT systems: timestamps, approvals, handoffs, exceptions. From those data points it reconstructs the process as it actually runs, in all its variants and workarounds.

The mirror is rarely flattering. Organisations discover their one designed process living dozens of parallel lives. What leadership believes about the operation meets what the event data shows, and the gap between the two is measured, in delays, rework and cost.

The market has moved with the need: Gartner now frames the field as process intelligence, platforms that combine mining, analytics, modelling and monitoring into one management layer over end-to-end flows.

Business intelligence answers what happened: which numbers moved, which targets slipped. Process intelligence answers how and why, because it keeps the sequence: which step produced the delay, which decision routed the case wrong, which handoff lost the customer.

The same traces read further. Object-centric mining follows the order, the delivery, the invoice and the complaint as they interweave, and journey-to-process analytics joins the customer’s experience data with the operational data behind each touchpoint.

That last connection makes the junction of the customer’s processes and ours visible: the moment of truth a customer feels, linked to the exact flow and step that produced it. Friction stops being an opinion and becomes an address in the process.

One layer, then, proves the article’s claim in practice. The process lives in the data, and the data gains meaning in the process. Davenport and Redman reach the same conclusion in Harvard Business Review: process management and AI reinforce each other, each making the other more valuable.

For the why, this layer is the proving ground. AI opportunities infused into strategic directions meet here the processes they must change and the data they must run on, while they are still opportunities, before they become AI use cases, prototypes and builds.

This is what stands between the why and the execution layer. Above it, ambition, strategic directions, processes and decisions. Below it, workflows, datasets and AI solutions. In between, the place where both sides meet.

One final test: the decision

Among everything the process levels map, one thing matters most here: they show where the business decides. The data members show what it decides with. A decision needs both at once. Preparing them in two separate audits prepares two halves of the same capability.

Most writing about decisions and AI starts at the operational level: which offer to show, which case to escalate, which route to take. Real, useful, and downstream.

The first decisions sit higher. What the business wants from AI, at which level of ambition, with what transformative intent: set where ambition fuses with strategy. Every operational decision inherits its direction from there.

The decision inventory looks in three directions. The decisions made today. The decisions the business does not know it should be making, invisible until someone maps where deciding belongs. And the decisions ahead, created by the ambition, each arriving with changes to prepare for.

So the test of readiness reads from the top down. Which decisions, today’s and the coming ones, will carry this ambition. Where in the processes do they live, or will they live. Does the data give them the meaning to act on. The decisions define which data is needed, never the reverse.

Value creation is the starting point. Processes and data are the connected lenses through which the business makes value creation understandable for AI.

The leap beneath the leap

AI ambition in AIVaaS™ draws one critical leap, between L2 and L3: below it, AI improves how the business operates today; above it, at L3 and L4, AI drives outcomes the business could not produce before. The leap is decided at the top. And a decision at the top does not land by itself.

This strategic leap has leaps beneath it, and this article has been preparing one of them: the leap of processes and data. Without it, the ambition above has nowhere to land.

Below this leap, AI improves isolated workflows. The gains are real: faster execution, fewer errors, lower cost. They are also temporary, and open to every competitor buying the same tools. Improvements like these settle into the new baseline of the industry.

Above it, everything assembled so far becomes one ground: processes known down to their decision points, and no longer only the internal ones. The outcomes above the leap, new offers, new ways of serving, live in the customer’s processes.

Data that carries meaning, one layer joining both worlds. On that ground AI begins to change how the business understands customers, makes decisions, creates value and operates end to end.

Orchestration is the nearest test. AI agents already work across processes, handing work to each other, deciding in sequence. They can cooperate only inside a shared frame, a constitution or a harness: the rules of the business, the boundaries of autonomy, the meaning of everything they read.

Here semantics stops being one member among three and becomes the precondition. One discount for a life moment is a single mistake. The same misreading, executed by a network of agents across every process, becomes the operating model.

Field evidence from 2026 agrees. BCG places the value of agentic AI in processes redesigned end to end, and names the technology and data foundation as the common blocker at scale. Deloitte counts roughly one organisation in five with a mature governance model for autonomous agents.

Above the leap something else opens as well. Customer needs read from data, including needs no customer has voiced yet, are where new offers and new business models begin. That is the next element of the third pillar, with its own article ahead.

The leap above is decided in ambition. The leap beneath decides whether it lands.

The ultimate why: Ana

The discount was never sent.

The agent had followed the workflow: silence, reactivation case, discount. The human at the decision point above followed the customer: eight years of loyalty and three days of silence deserved a voice. Call her.

The agent did not sell. It listened. And Ana told it what no system could have known: three days earlier, a test had shown two lines.

Congratulations, said the agent. No discount, no recommendation. It noted a life moment, paused the automated nudges, set its tone to warm and patient. Presence, executed by a machine, decided by the business.

Then Ana said the sentence every executive waits for: she will probably start buying completely different things now, she does not yet know which.

A need no customer has voiced, forming in real time. The new offers above the leap, and the value they carry, begin exactly here, in a customer’s process the business was present for.

Readiness made the difference, end to end. The why, set in ambition, gave the direction. The process knew where the decision lived. The data caught the signal. The layer in between connected them. And the agent acted on the meaning it was given.

One link in that chain stays human. The stories behind this series call it the capacity to be surprised: a surprise turned into an algorithm becomes an expectation. Readiness for it means keeping people at the decision points where the unexpected arrives.

One day her child will be a customer, with data in the systems from the very first day. Serving that customer will raise the question this article started with, at full weight: ready for what?

Ana answers it better than any framework. Ready for the customer: for her processes, her moments, her needs before she can name them. Everything else is the execution layer.

The AI Value Boardroom builds the AIVaaS™ methodology article by article. The next element opens where this one closed: customer needs, read from data, becoming new offers and new business models. Subscribe free to follow the arc from ambition to the ground it lands on.

In AIVaaS™, AI governance is not used only in the classical sense. It means the leadership mindset and approach that determine whether AI adoption becomes real business transformation.

The previous article in this series, AI Transformation and the Human Veto, ended where every serious AI transformation sooner or later arrives: at a question of leadership.

This time, though, it came from the employee’s side. The question is not whether the AI tool is useful. It is whether the future being built still has a place for them.

This is where AI governance begins.

Not just another committee. Not a policy archive. Not a control layer added late, once AI initiatives have already begun to spread across the organisation. Those elements may be necessary, but they are not the centre of the matter.

In AIVaaS™, AI governance is the leadership function that decides what the organisation is trying to become with AI, how far AI may go in creating value, who owns the critical calls, whether the vision for AI keeps a place for the people who do the work, and how the AI transformation stays coherent while business strategy, people, processes, and technology all move at once.

That makes governance a leadership decision before it is a control system.

Why the usual view of AI governance is not enough

The classical governance view has a legitimate role. It has to address accountability, transparency, security, privacy, data quality, model behaviour, human oversight, escalation, monitoring, and compliance with the relevant regulation.

As AI becomes more autonomous, that question grows sharper still, because what matters is no longer only what AI says, but also what AI does.

So this part should not be dismissed. It is necessary. The problem is that it is not sufficient.

Why?

Because by the time classical governance gets to work, the larger decision has usually already been made, often without anyone naming it. At that point the organisation has already decided whether AI will mainly support existing work, improve the operating model, reshape the value proposition, or become part of the business strategy itself.

That earlier decision is the deeper governance act.

An organisation can have policies, review boards, model registers, and approval flows, and still govern AI poorly. It can control AI activity while never asking whether that activity belongs to a coherent ambition for both AI and the business. That coherence is what AIVaaS™ calls AI Ambition Fusion. Traditional governance can make AI safer. But does safer AI matter for the business transformation?

This is why AIVaaS™ places governance in the Foundation. It gives every act of AI adoption a frame, and asks not only whether it is allowed, but whether it belongs to the ambition and vision the business has chosen. In that sense, AIVaaS™ AI governance protects coherence.

What AI governance means in AIVaaS™

The AIVaaS™ view of AI governance does not replace the classical one. It absorbs it into a broader transformation logic.

AI still has to be used responsibly, so risk, ethics, compliance, and oversight remain necessary. However, the governance of AI must address a much more important question: what must be governed to ensure the successful adoption of AI as a business transformation?

That is why AIVaaS™ governance has six connected elements.

First, the organisation of governance, which begins with accountability. AI governance is not a separate committee. It sits with leadership in the boardroom. The highest AI governance body is the board itself, extended with the execution roles that make transformation real, the Chief Transformation Officer, the Head of the Strategic Project Office, the Head of Enterprise Architecture, the Lead Business Analyst, and others. Its strategic side decides where AI should take the business and how far the ambition goes. Its execution side turns that ambition into work, capabilities, decisions, and progress.

Second, purpose, which sits deeper than direction. Direction is where the organisation is going; purpose is why it exists at all. Governance has to be clear how AI serves that purpose and the value the organisation owes the people it serves. When purpose is vague, AI activity is easy to launch and hard to judge, and pilots and productivity stories pile up without telling anyone whether the work serves what the organisation is for.

Third, AI ambition and AI vision. AIVaaS™ does not treat the AI vision as a separate technology document beside the business strategy. The vision comes from AI ambition and is fused into the strategy itself, so the organisation does not end up with two worlds, an official business strategy and an unofficial AI agenda looking for room inside it.

Fourth, readiness, captured as Business AI-Ready. Successful adoption is not produced by a tool or a single capability. AIVaaS™ states it as a formula: Business AI-Ready combines Business-Ready, Organisation-Ready, AI-Ready, and Implementation-Ready. Governance has to hold those four states together, because transformation fails when the parts move at different speeds and nobody owns the integration.

Fifth, compliance, ethics, and risk. This is the classical part, and in AIVaaS™ it is one part of the apparatus, not the whole of it. It matters because AI must be trustworthy and controlled, and it becomes too narrow when it defines governance as protection alone.

Sixth, decision ownership and escalation. This is the most neglected part and one of the most important. Governance has to define who has the final word, which criteria guide the call, what is escalated, and when a decision moves from operational management to leadership. Without it, an organisation can look governed while the real decisions stay scattered across functions, projects, and personal interpretation.

Together, these six elements change what governance means. It is no longer a set of constraints around AI use. It becomes the leadership mechanism that keeps AI ambition, business value, organisational readiness, and responsible use connected.

That is why the first governance decision is not a policy. It is ambition.

The first act of AIVaaS™ Governance is authorising ambition

AIVaaS™ introduced four levels of AI ambition earlier in the series. L1 supports people in their existing work. L2 improves the operating model. L3 adds new business models and sharpens differentiation. L4 runs AI through the whole business strategy and disrupts markets.

The critical line sits between L2 and L3.

Below that line, AI mostly improves what already exists. Work becomes faster, cheaper, more consistent, and easier to manage. This matters, because efficiency is a real form of value, and many organisations need that operational discipline before they can move further.

But it is not yet transformation.

Above the line, AI begins to change what the organisation can offer, how value is created, how customers are served, how decisions are made, and what competitors must respond to. The logic moves from better operations to new forms of value.

This leap cannot be delegated to a technology function, and it cannot be left to a portfolio of use cases. It has to be authorised by leadership, because it changes the organisation’s understanding of value, risk, investment, capability, and future advantage.

This does not mean reaching L3 or L4 immediately. Delivery can be gradual, and capabilities, processes, data, roles, governance, and literacy can mature over time.

But the destination has to be authorised early in AI governance.

If the boardroom communicates only an L2 story, the organisation behaves accordingly. Managers search for productivity gains, finance looks for cost reduction, employees hear the threat, and customers see little difference. AI becomes active in many places and transformative in none.

Why managers stop at Level 2

The difference between a manager and an AI leader is not the job title. It is the level of value they look for.

A manager usually sees AI through the operating model. The question is how to make existing work faster, cheaper, more accurate, or easier to control. AI becomes a tool inside the current organisation, placed into current processes, current measures, and current assumptions about how the business works.

This view is completely legitimate, and in many organisations the first responsible one, because operational reality cannot be ignored. L2 can remove friction, improve quality, increase consistency, and create the capacity for the next move.

But the managerial view has a ceiling.

When AI is understood mainly as an operational tool, governance becomes a matter of permission and control. Which tools may be used. Which data may be entered. Which outputs must be reviewed. Which risks must be prevented. The organisation becomes more controlled, but not more valuable to the market.

And the same operational gains are open to everyone. Competitors reach L1 and L2 as well, so efficiency that every rival can copy does not create lasting advantage. AI can make the organisation leaner and safer and still leave its position in the market unchanged. Real transformation asks for more.

The AI leader sees a different horizon.

For the AI leader, AI is not primarily a tool inside the existing business. It is a business power that can change the value the organisation creates. The question is not only where AI can improve the current process. It is what becomes possible for the customer, the employee, the partner, and the owner when AI changes the economics of the work.

This is the move from management to leadership.

Management asks how AI improves what the organisation does.

Leadership asks how far AI should transform the business.

That is a question of vision, and a vision too small to require change will not produce it.

This is also why AI governance cannot be reduced to a delegated function. A CAIO, CDO, CIO, or transformation office can build the machinery. They can coordinate, advise, enable, and execute. But they cannot replace the leadership decision about ambition.

Leaders cannot stand outside the transformation and sponsor it from a distance. The decision about how far AI may go is theirs to make.

Governance without leadership becomes administration.

Leadership without governance becomes aspiration.

AI transformation as business transformation needs both.

Culture is the human in the loop of AI transformation

The culture an organisation has does not stand on its own. It follows from its strategic posture, the character the organisation has already chosen. Strategic Posture Readiness set that character out earlier in the series. The point here is what it does to culture.

A Defender’s instinct produces a culture of control and hierarchy, and a transactional style of leadership that runs on targets, rules, and reward for compliance. A Prospector’s produces a culture of experiment, close to what the competing-values research calls an adhocracy, and a transformational style that works through vision, trust, and the readiness to change the business itself. An Analyzer sits between them, and its leadership has to hold both at once.

None of this is a law, and real organisations are mixtures. But the tendency is strong enough to matter, because culture is where a higher ambition is either allowed to operate or quietly pulled back. If the ambition says L3 while the culture rewards L2 behaviour, the gap is not motivational. It is structural, and governance has to notice it.

This is why culture, more than the org chart, is the human in the loop of AI transformation.

In AI orchestration, human in the loop describes where people review, approve, or step into what the machine does. The useful lesson is not that a human should be inserted somewhere into the process. It is that human judgement matters only when the human has the context, the authority, and the capability to improve the decision.

A reviewer who does not understand the output is not governance. A reviewer without authority is not governance. A review step that records approval but never disagreement is not governance. It can look like control, but it does not produce judgement.

The same logic applies to the introduction of AI into the business. Leadership and the culture it builds decide what the transformation means, how far it may go, and whether it becomes trusted enough to move.

A culture of trust and experiment is therefore not a soft addition. It is part of the operating conditions for higher AI ambition.

Employees must see that AI is not being introduced only to extract savings from their work. Managers must see that their role is not to protect the old model with new tools. Leaders must show that the transformation is not an efficiency programme disguised as progress.

This is where the human veto returns. Employees do not resist value. They resist a value story in which they carry the cost and someone else receives the gain.

Governance is what keeps that story from becoming the default. Trust does not come from saying AI will help people. It comes when the governance of AI makes that true.

AI literacy is judgement, not tool training

AIVaaS™ places AI literacy in the second pillar, as a readiness of its own. Like the shift from manager to leader, it works at two levels, and only one of them carries transformation.

Many organisations treat AI literacy as a training programme. People learn what generative AI is, how to write prompts, what not to enter into a tool, and how to use approved systems responsibly. This is useful, and for L1 and L2 it is often enough, especially when AI is spreading quickly across everyday work.

But it is too small for leadership.

In AIVaaS™, AI literacy means understanding AI as a business power. It means grasping how AI changes the economy, customer behaviour, competition, operating logic, decision making, and value creation. It means seeing the difference between improving a task and changing a business constraint.

This deeper literacy does not sit on its own. It is closest to AI governance, which sets the ambition it has to understand, and it shapes leadership and culture, because people cannot lead or trust a transformation they cannot read. That is why it belongs in the second pillar while drawing its force from the Foundation.

Employees, in turn, have to know where AI supports them, where it reshapes their role, and where their judgement becomes worth more, not less.

Without this, governance turns formal. People follow rules they do not understand, reviewers approve outputs they cannot weigh, and managers favour use cases that look efficient but do not move the ambition. Leadership mistakes visible AI activity for transformation.

Literacy makes governance real because it lets people judge. The human in the loop has to be able to think.

The manager has to separate efficiency from value. The leadership team has to see whether the portfolio is still trapped at L2 while the market moves to L3.

That is not tool skill. It is strategic literacy for the AI age.

Ana is the final test of governance

There is one test that decides whether AI governance is really working.

Look at Ana.

Not only the process, not only the internal business case, not only the governance dashboard. Look at the customer the organisation exists to serve.

If governance produces a safer organisation that customers cannot feel, it has done only part of the job. If it makes operations more efficient without changing the value Ana receives, it has stayed below the line.

If it produces approvals, policies, and committees but never asks what becomes better for Ana, it has governed AI activity, not AI transformation.

This matters more in the intelligent economy, because Ana is changing too. She no longer chooses on her own.

Increasingly she comes with an agent that compares, filters, remembers, and decides with a discipline ordinary persuasion cannot easily overcome. It is not impressed that a process was automated. It cares whether the value is better, clearer, faster, and more trustworthy.

Ana’s agent raises the standard, and that is why L3 and L4 matter. The customer side of value is getting harder to fake.

Ana also gives employees a reason to believe. When the organisation can show AI is being introduced to create more value for the customer, and not only to cut cost, the transformation gains a different logic.

Through Ana, AI governance, leadership, culture, and AI literacy meet in one place. Governance decides whether AI is allowed to create higher value. Leadership holds that ambition. Culture makes it believable. Literacy makes it understandable. And Ana tests whether it matters.

Four articles in, the series has turned AI pressure into value. The opening diagnosis named the failure. The next two read the economy from the outside and tested the organisation’s character, and AI Value Has an Address gave every opportunity a stakeholder it serves.

That portfolio completes the first AIVaaS™ pillar, AI business value definition. It is value on paper. Nothing on paper moves a customer, and nothing on paper survives contact with the employees who have to carry it.

So the methodology enters its second pillar, organisation enablement. The previous article left one question open: how each stakeholder perceives what is being planned for them. That question now decides everything.

And a person can decline. A portfolio that is net positive on paper still fails when one stakeholder reads it as a threat and quietly withholds the work.

Every AI opportunity the organisation priced carries a risk it did not price. The AI transformation has to be carried out and absorbed by employees, and any one of them can veto it.

The veto is a feature of ordinary organisations, where AI transformation arrives from above and employees are simply expected to accept it. In a young technology firm, however, employees adopt AI of their own accord, so the veto rarely arises. Most organisations, though, are not like this.

Where the veto begins

Distributing value across stakeholders was the last article’s work. It is necessary, and it is not enough. What decides whether the change is accepted is not the spread on paper, but whether each stakeholder sees the value as meant for them too, and believes it will really reach them.

So, every opportunity has to be read through the stakeholders it touches, one at a time, a person, a team, or a whole department.

Four questions decide how the value lands. What the stakeholder gains. What they pay or risk, in time, control, or the job itself. How they see it, not how the plan assumes they should. And whether their support is needed for the opportunity to work at all.

One of the four carries the most weight, and it is the one most often skipped: how they see it. Perception is not what the value is. It is what they believe it is, and belief is what they act on.

The same opportunity often reads two ways at once. For example, shorter handling time is a gift to the customer and a warning to the employee who runs the process, because the time saved is the work removed. One opportunity, two perceptions, and the veto sits with the second.

This is why the stakeholder work is more than analysis. To name who an opportunity serves, and to ask how they see it, is the first act of involving them in the change. Value defined with people, rather than for them, is far harder to veto.

Underneath perception sits something slower to build: trust. How far do employees believe leadership, and the way it says the value will be shared. A plan can be sound and still be disbelieved.

Trust is built from memory. Past experience shapes it, and so does any disappointment already lived through. When earlier initiatives promised much and returned little to the employees who delivered them, the next promise lands on that record. Each unkept one makes the veto more likely.

And where employees see the value flowing mainly to the organisation, its leaders, and its owners, the veto is already forming. Not because the value is false, but because it is not theirs.

How the veto looks in practice

The veto rarely shows itself at the start. Employees often welcome AI when it first appears, because the early version lifts a tedious task off the desk. The relief is real. It is also small, and it does not last.

The fear arrives later, when employees see the rest of what the tool can do. Not the one task it took off the desk, but how much of their work it can carry, in a fraction of the time and often at higher quality.

That is the turning point. The same capability that saved an hour could, scaled across the role, reshape it or remove it. Maslow’s hierarchy is blunt about this: while survival is in question, the higher rewards wait.

What employees settle into is a quiet verdict. The value of this AI is the company’s, the saving belongs to the owner, the risk belongs to me. That verdict is the veto, and from where they stand it is rational.

The ground was thin before AI arrived. Gallup’s State of the Global Workplace 2026 puts engagement near one in five worldwide, and finds most who use AI report a productivity gain while almost one in five expect their role cut within five years.

The last number is that fear in the data. The employees who gain most from the tool are the same ones who expect it to take their role, and the fear of being replaced outweighs the gain.

A narrow value is not only a failure of communication. It is a tell about the AI Ambition behind the plan. An organisation that points AI at cost and efficiency alone, at L1 and L2, has nothing larger to offer, because it set out to save, not to build.

Behind that ambition is the organisation’s strategic posture, its character: how it reads the world outside and inside, and where it therefore chooses to create value.

A Defender posture turns inward by instinct. It protects what exists, and rarely reaches for the AI opportunities that open new value for customers or break new ground in the market. Those are the very ones that give employees a reason to move.

The posture gap becomes a motivation gap. The trap is conditional, and the condition is the opening. The arc from welcome to veto runs only where AI is introduced without a story, as a saving with no shared future attached.

Making the value theirs

Lifting the veto is not a campaign to convince employees that a one-sided deal is fair. It is the work of making the value theirs, and it runs on three moves held together: a vision from leadership, change managed through value, and the employee experience and engagement that follow.

The first move is leadership, and it leads for a reason. Employees read a change by who stands behind it, and one the leadership does not visibly own is one they can quietly refuse.

Leadership here is a vision before it is a plan. People accept the change when the organisation’s bright future is visibly their bright future too, and not a future where the gains are the company’s and the costs are theirs.

That vision carries a commitment employees can hold the organisation to. The promise is a balanced fusion of employees and AI agents, built in their favour, not against them.

The agents take over the routine that drains the day, and the employees are freed for the creative, higher work the machines cannot do. The judgement stays with the employee, and there is more of this better work ahead, not less.

What this asks of leaders is the subject of the next article. Here it is enough that the promise is credible, because a vision employees do not believe is one more reason to veto.

The second move is change management, run through value instead of through steps. The familiar models move employees along a sequence. AIVaaS™ begins from another question: what each employee stands to gain, and whether the plan lets them share in it.

The promise of a share in the value is matched by a real shift in the daily experience of work. Not lighter tasks. That is the small relief that already failed to convince. Something higher than that.

More success at the work that matters, as employees reach the goals and the mission they are there for, achieving more of it rather than doing less. And greater personalisation, as AI shapes the tools and the workplace around how each employee works best.

More confidence in the right decisions, with AI as support rather than threat. And better wellbeing, as the routine that consumed the most time falls away.

And some of the reward comes from outside the company. When the value reaches customers, their thanks and their praise return to the employees who built it, and recognition of that kind moves motivation more than any internal measure.

These are not perks. They are the conditions under which trust recovers, the veto lifts, and engagement and motivation rise, because employees now see value that is theirs.

And only then can the organisation reach the more far-reaching forms of AI. The orchestration of agents, where they take over more and more of the employees’ own work, asks for even more trust than the first steps did.

They accept it only when they already hold the end vision, have had their fear answered, and find direct and indirect value in it for themselves. Where that holds, the veto is behind them. Readiness is not a technical state, it is a willing one.

Why this matters for Ana

In the stories that open The AI Value Boardroom, Ana is the customer the organisation serves today, and the child she carries is the customer of the next, more intelligent economy.

Ana never meets the portfolio, the planning, or the veto. She meets a person, or a service a person built, and she feels whether someone cared. With a birth approaching, she reads that care more closely than before.

Care for each customer comes from the experience and the engagement of the employee behind it. This is why the employee experience (EX) and the customer experience (CX) are two sides of one coin.

People who have quietly vetoed the change deliver Ana a thinner version of everything planned for her. People who believe in it deliver what was promised. The value reaches her through people who chose to carry it, or it does not reach her at all.

The series so far has moved in three steps. The opening diagnosis showed that most AI initiatives still fail to produce measurable value, and that the missing piece is a methodology for turning ambition into stakeholder value.

Outside-In AI Impact read the economy from the outside and derived the level of AI ambition the environment demands. AI Ambition Needs the Right Character tested whether the organisation’s strategic character can carry it.

An ambition that clears both readings is credible. It has survived the environment, and it has survived the mirror. Credible, however, is not the same as productive. A credible ambition still produces nothing by itself.

Where do concrete AI opportunities come from. Which ones deserve investment first. By what measure does the organisation know they were worth it. The last two elements of the first AIVaaS™ pillar answer the first two. One function from the Foundation underneath them answers the third.

The race moved to value

Across the major research houses, the 2026 reports converge on one message, stated in different words. The conversation is no longer about whether organisations use AI. It is about whether they can turn AI into value.

PwC finds that a fifth of companies generate nearly three quarters of all financial returns from AI. The race is not decided by who adopts. It is decided by who converts.

Capgemini describes ROI measures moving beyond savings, toward revenue growth, risk management, and customer experience. Efficiency itself is becoming a commodity. Organisations that stay with efficiency-only AI hit a ceiling, because the basic savings run out.

AIVaaS™ draws one structural conclusion. If value is the race, the measure of value cannot be an afterthought. It has to be committed before any execution begins.

That is why Value & ROI Validation sits in the Foundation of the methodology, beneath all three pillars. The commitment is made up front: realized business value will be measured, not adoption. Continuously, not as a milestone at the end.

With the measure committed, the two remaining elements of the first pillar can do their work.

Value has an address

Value is relational. Something has value for someone, in some context. Stated like that, it sounds obvious, and yet most AI opportunity lists violate it in the first column. They name a technology and a process, and leave the recipient implicit.

An AI opportunity that cannot name the stakeholder it creates value for is not a business opportunity yet. It is a business wish.

It also carries a practical penalty downstream: its KPI floats. Shorter handling time is value for a customer. Lower cost per transaction is value for an owner. Less repetitive work is value for an employee. A KPI without a recipient measures nothing in particular.

An opportunity without a firm KPI falls out of every serious prioritization, because nobody can say how much it brings, and to whom.

AIVaaS™ therefore reverses the usual order of the question. Not where could AI help, but for whom should value be created, and what kind. The search runs in two directions at once, and both are mandatory.

Outward, toward customers, partners and the market: faster and more personal experience, offers that were not possible before, new sales opportunities, better connected value chains. Value in this direction shows up as revenue growth and differentiation, the difference a market notices.

Inward, toward employees, leaders and owners: a better working experience, fewer repetitive tasks, faster and better supported decisions, healthier returns. Value in this direction shows up as productivity, lower cost, and better decision quality.

Organisations that search only inward end up with AI operations. Efficient, measurable, and quickly forgotten by the market. Organisations that search only outward meet resistance, because the people who must deliver the change see no value in it for themselves.

The portfolio that works covers both directions, by design.

The series introduced four levels of AI ambition earlier, from L1, where AI supports people in their existing work, to L4, where AI is the core of the business model.

The red line on the arrow is the critical leap, the same boundary the series marked between L2 and L3. The forms to its left are within reach of L1 and L2, where AI improves existing work. Everything to its right opens only with an L3 or L4 ambition.

The stakeholder profile of the opportunities also performs a quiet diagnostic. It reveals the organisation’s actual ambition, regardless of the declared one.

Opportunities that address mostly internal stakeholders place the real ambition at L1 or L2, whatever the strategy slide says. Opportunities that reach external stakeholders in ways that were not possible before move the organisation toward L3 or L4.

A stakeholder-unbalanced portfolio is an early signal that declared ambition is higher than actual.

The stakeholder defines the value. The value defines the measure. The measure makes prioritization possible.

A portfolio, not a shopping list

Prioritization in AIVaaS™ is not the act of picking the most attractive individual ideas. It is a strategic decision about distribution, derived from the ambition the previous readings established.

The wrong question is which quadrant to invest in. The right question is what distribution across quadrants fits the organisation’s strategic position.

The quadrants come from the Value Creation Matrix. Its vertical axis measures the scope of change, from the operational model, how the organisation works, to the business model, what value it creates. Its horizontal axis measures value capture, from internal value to external value.

Four quadrants emerge. Q1, cost savings, the hygiene factor that funds the rest. Q2, revenue growth and enhanced customer value, the growth engine. Q3, the radical productivity leap, where breakthrough begins on the inside. Q4, new value forms, where differentiation and disruption live.

If the AI Ambition Matrix earlier in the series was the compass, this matrix is the X-ray. It reveals hidden value because it changes the unit of analysis.

A task focus produces Q1: automate what exists, optimise locally. A constraint focus opens Q3 and Q4: find where AI removes a critical business constraint, the one that kept a segment unserved or an economics unworkable. New value appears where the task list showed nothing.

The balance across the four quadrants is an allocation decision. As a starting point inside the methodology, not a dogma: roughly 35 to 45 percent in Q1, funding the rest, 25 to 35 in Q2, 15 to 25 in Q3, and 10 to 20 in Q4, adjusted to the organisation’s position.

The warning signs matter more than the percentages. A portfolio that sits in Q1 alone does not survive long term, because everyone is harvesting the same savings. An empty Q4, where the environment demands transformation, leaves the door open to whoever fills it first.

The matrix is a tool inside a step where business direction is already clear. It distributes a committed ambition. It does not substitute for one.

The two matrices stay distinct, and they connect. Ambition levels accumulate, because a higher level keeps doing everything the lower ones do. So ambition sets the reach of the portfolio: L1 reaches Q1, L2 adds Q2, L3 opens Q3, L4 opens Q4.

An organisation with a credible L3 ambition lives across Q1 to Q3.

Two prioritizations, one trap

Prioritization happens twice in the methodology, not once, because early on feasibility is unknown. Process and data maturity, and the readiness of people, only become measurable later.

The early prioritization, the one described here, measures value alone. It produces a shortlist of opportunities worth deeper analysis.

The portfolio is also a strategic checkpoint, the last one where the selection can still be steered. Every opportunity that survives must belong to the business strategy, because AIVaaS™ knows no separate AI strategy to retreat to.

This is where the organisation verifies Ambition Fusion in practice: whether the AI ambition and the business ambition are still one and the same. A portfolio that drifts away from the strategy has already given the answer.

Value alone never meant size alone. Every opportunity on the shortlist carries two more readings: who receives the value, and whether it serves the ambition the organisation committed to.

The trap looks rational at every step. An organisation commits to L3, where AI drives new offerings. Then the portfolio fills with Q1 savings, because those score easiest, and each one is approved on its own merits.

A year later the portfolio says L1. Nobody voted against the ambition. The budget did, one approval at a time.

The drift has a second engine: power. The preference for measurable savings tends to sit with whoever approves the spending, and a tension that nobody governs is resolved by the strongest preference in the room.

That is why AIVaaS™ treats priority-setting as governance, not as a scoring exercise. The Foundation commits the rules for how priorities are set up front, next to the measure of value.

The final prioritization comes later, when opportunities convert into concrete AI use cases, with actors, steps, data and rules. There it weighs value against feasibility. Two different moments. Two different objects.

The measure that holds the portfolio

A portfolio is only as strong as the measure underneath it. Bare ROI, used as the sole criterion, has a systematic bias. It computes most easily where value is a direct saving on something that already exists.

An ROI ranking therefore always favours the left side of the matrix. Q4 measures the economics of something that does not exist yet. There is no baseline to compute against, so in a bare ROI table it loses every time. And it is where the most durable value lives.

The bias is not only ROI’s. Execution KPIs lean the same way, toward the left of the value arrow, because targets are easiest to define there and because most KPIs read in quarters, not years.

Customer-side value resists simple measures. NPS is one of the few that holds, and one instrument cannot carry half the arrow.

The boundary is the same critical leap the series crossed before. In ambition it separates L2 from L3. In value creation it separates the operational row from the business model row. In measurement it separates what bare ROI and standard KPIs can read from what they cannot.

The conclusion is not that ROI is wrong. It is the right instrument for the left quadrants, where value arrives as direct savings or direct revenue on the existing model.

Further right and further up, the measure has to widen: new capabilities, customer and employee experience, market position, the economics of a segment that could not be served before.

This is the work of Value & ROI Validation, the commitment from the opening, running through the transformation rather than waiting at the end of it. Every opportunity on the shortlist, and every investment that follows, is held to one continuous question.

Is this creating value, for whom, and how does the organisation know. The address is what makes the answer checkable.

Why this matters for Ana

In the stories that opened this series, Ana is the customer the organisation serves today, and the child she carries is the customer of the next, more intelligent economy.

Prioritizing by value is the act of choosing what to build for Ana first. Ana never sees the portfolio. She never sees the quadrants, the allocation, or the shortlist. She sees only what reaches her.

A portfolio that looks balanced on paper but leans toward internal savings produces nothing she will ever notice. The address on the value decides whether she ever receives it.

The series moves next to the people who have to deliver all of this, and to a question left open here: how each stakeholder perceives what is being planned for them. Engagement and motivation come ahead of everything else there.

In the previous article, Outside-In AI Impact, AI ambition was read from the outside. The economy moves, customers and competitors are reshaped by AI, and the organisation derives the level of ambition that this outside pressure demands. That reading produced the AI Ambition Matrix, a level of ambition committed to inwardly. If the framework is new, the link is here, because this article picks up exactly where it left off.

But that reading left a question open. The level an organisation chooses and the pressure it actually faces are not always aligned. When they are not, the gap becomes the central strategic problem. It was named there. It was not yet closed.

This article closes it. And the closing happens inward, because the gap has two layers, and both of them live inside the organisation rather than out in the market.

The first layer is whether the organisation aimed high enough for the environment it is actually in. The second is whether it is, by character, capable of reaching the level it aimed for. Most organisations measure neither. They measure ambition, write it down, and move to execution. The ambition stays on paper because nobody asked the two questions that decide whether it can leave the page.

The economy is not a fixed picture

Where the reading began, the nature of the economy was the subject: industrial, digital, intelligent, each rewarding a different logic. That description was a snapshot. What it did not capture is motion. How fast the ground is moving under the organisation, and how predictably.

This is where Igor Ansoff still matters, decades after he wrote. Ansoff, often called the father of strategic management, spent twenty-five years studying why firms succeed or fail in turbulent environments. He placed environments on a scale of turbulence, from stable and repetitive at one end to surprising and discontinuous at the other.

Turbulence is not a mood word for difficult times. It is a measurable thing, built from a few concrete questions. How fast are customer needs changing, and how predictably. How often do new competitors arrive, and how often from outside the industry entirely. How quickly do products and business models age. How much of the future can be read from the past. A stable environment answers all of these with slow, predictable, rarely, rarely, most of it. A turbulent one answers with fast, discontinuously, constantly, often from nowhere, and almost none of it. The point is not the exact level. The point is that an organisation built for one set of answers cannot survive a different set.

From this, Ansoff produced a single rule that survived empirical testing across more than a thousand firms.

For optimum profitability, the levels of both the strategic aggressiveness and the general management responsiveness of the firm must be aligned with the environmental turbulence level. Aggressiveness is how boldly the organisation behaves. Responsiveness is whether its capabilities can keep up. Both must match the turbulence outside.

This is not a soft rule. Ansoff observed that as the gap between the organisation and its environment widens, performance falls faster in intensely competitive markets than in calmer ones. And it does not stop at decline. The organisation that stays matched to an environment that has moved on does not simply earn less. It loses the right to compete at all. Match the turbulence and the organisation lives. Fall behind it and the organisation is removed, not by a single dramatic failure, but by a market that has quietly stopped needing what it was built to do. This is the rule underneath everything that follows.

Ansoff built this into three propositions. Strategic aggressiveness must match environmental turbulence. Internal capabilities must match strategic aggressiveness. And the organisation’s subsystems must operate in harmony with each other. That third proposition is worth holding on to. It returns at the end of this article, and it is the part almost everyone forgets.

AI is raising the turbulence

Here is what changes the picture now. AI is not only a technology operating inside the economy. AI is raising the turbulence of the economy itself, and it does so along the very dimensions the opening of this article used to define turbulence.

Customer needs change faster, because customers recalibrate their expectations against every AI-enabled experience they meet elsewhere. Competitors arrive more often and from further away, because AI capability crosses industry boundaries more easily than any physical advantage ever did. Products and business models age more quickly, because an AI-native offering can make an established one feel dated within a single cycle. And the future becomes harder to read from the past, because the patterns that planning relied on are exactly what AI is rewriting. Each of these was already a source of turbulence. AI pushes every one of them upward at the same time. Agentic systems continuously redefine how enterprises operate while forcing them to rethink how they create value, and the pace of that change keeps rising rather than settling.

Put Ansoff and this condition together and the consequence is sharp. AI lifts the effective turbulence of the environment upward, and it does so continuously, not once. An organisation that was well matched to its environment a year ago can find itself behind without having changed anything at all. It did not get worse. The ground moved. The previous alignment quietly became the new misalignment.

This is the first layer of the gap, and it is the quietest, because nothing inside the organisation signals it. No project failed. No number dropped on the day it mattered. The organisation kept doing what used to fit, while the environment it was fitting to dissolved underneath it. An ambition set for last year’s turbulence is already too low for this year’s, and there is no internal alarm for that.

Who the organisation is, not what it aims to be

Turbulence tells the organisation how high it needs to aim. It does not tell it whether it can reach. For that, ambition is the wrong instrument entirely, because ambition describes where the organisation aims to be. The thing that decides whether it gets there is something slower and harder to change: who the organisation is.

Ansoff had a name for this. He called it the organisation’s strategic posture: its fundamental relationship to uncertainty, competition and growth, the underlying stance that shapes every strategic choice downstream, from where it invests to how it competes to what it measures. Posture is not one decision among many. It is the logic that sits above all the decisions. And it is a whole-pattern thing, the organisation’s relationship both to the environment outside and to itself inside, which is why it is so easy to leave undefined and so dangerous to leave that way.

Because here is what happens when posture stays implicit. Leadership declares an ambitious vision, the fused ambition of the previous article, an innovator’s posture on paper, while the systems underneath it, the KPIs, the capital allocation, the structure, the way managers are rewarded, all quietly reward defensive behaviour. The organisation says one thing and is built to do another. This is structural hypocrisy, and it is the real reason so many strategies never survive execution. The strategy is not wrong. The posture was never made explicit and built into the machine that has to carry it.

Most organisations, in other words, do not have a strategy problem. They have a posture they never named. The discipline is simple to state and hard to do: before defining where the organisation is going, establish who it is, and if there is a gap between the two, that gap is the first strategic priority, ahead of everything else.

To make posture concrete, it helps to give it types. Miles and Snow, building on the same Ansoffian foundation, named three viable forms of strategic orientation. This is the working language for the rest of the article: posture is the concept, orientation is its visible character, and there are three of them that hold together.

There is a fourth pattern, but it is not a strategy. The Reactor is what appears when an organisation solves its problems inconsistently, declaring one thing and rewarding another. The Reactor is not a fourth way to compete. It is a warning light. It is structural hypocrisy showing up as a type: the organisation’s subsystems out of harmony with each other, which is exactly the third Ansoff proposition failing in plain sight.

None of these orientations is better than the others. A Defender in a stable market is perfectly placed, and a Prospector there is burning money. The character is only right or wrong relative to the ambition it has to carry and the turbulence it has to survive.

The posture gap

Now both instruments are on the table. Ambition says where the organisation wants to go. Orientation says who it is today. Lay one across the other and the second layer of the gap appears.

On the diagonal, the cells resonate. A Defender pursuing a defensive ambition, where AI optimises how the organisation already operates, is aligned. A Prospector pursuing market disruption is aligned. Off the diagonal, the gap opens. A Defender reaching for disruption is the most dangerous cell on the grid. The ambition demands continuous reinvention, high risk tolerance, decentralised decision-making. The character supplies cost discipline, low risk, central control. The organisation is not lazy and not wrong about where the market is going. It is structurally unable to get there with the character it currently has. The ambition is real on paper and impossible in the building.

This is where the AI ambition levels meet a hard threshold. L1 and L2 keep AI inside existing work. AI supports people, or it augments existing processes, but the operating model and the business model stay as they are. A Defender or an Analyzer can carry that. The critical leap is between L2 and L3. At L3 and above, AI begins to drive outcomes the organisation could not produce before: new offerings, new business models, a repositioning in the market rather than an optimisation inside it. That leap is not technological. It is a leap in character. The table below maps each level to the orientation it requires.

Below the leap, almost any orientation can cope, because AI is only making the existing organisation faster. Above it, the Defender orientation runs out of room. An organisation reaching for L3 or L4 with a Defender character is attempting the most common and most expensive failure in AI transformation: trying to jump from operational excellence to strategic transformation without the disposition that strategic transformation requires.

The gap between where the organisation wants to go and who it is today is its strategic posture gap. Its value is not that it judges. Its value is that it shows what has to change first, before any AI project is worth starting. Close the posture gap and the ambition becomes credible. Leave it open and every initiative downstream inherits the misalignment.

Two layers, one reading

So the gap that the previous article left open is not one gap. It is two, stacked. The first layer asks whether the organisation aimed high enough for the environment it is actually in, the environment whose turbulence AI keeps raising. Aim at L2 in an environment that now demands L3, and the misalignment is there before character is even considered.

The second layer asks whether the organisation’s character can carry the level it aimed for. Reach for L4 as a Defender, and the ambition cannot survive contact with the organisation that has to deliver it.

A credible ambition has to clear both. High enough for the turbulence. Matched to the character that has to execute it. An ambition that clears only the first is a wish exposed to the market. An ambition that clears only the second is comfortable and slowly fatal. Only when both layers close does ambition stop being paper and become something the organisation can actually do.

Closing the gap is not the same as lowering the ambition to fit the character. That is the wrong move, because the turbulence does not lower with it. The gap can also close from the other side: by reshaping the character to carry the ambition the environment demands. Character feels permanent, but it is not. It is built into systems, incentives and leadership behaviour, and what is built can be rebuilt.

Why this matters for Ana

In the stories that opened this series, Ana is the customer the organisation serves today, and the child she carries is the customer of the next economy, more intelligent still.

Reading Ana from the outside was the whole point before, because the shift around her runs through every dimension of the economy. This article adds the harder half. Reading Ana is not enough. An organisation can see exactly what she will need, and exactly what her child will need, and still fail to deliver any of it, because the organisation that has to build it is the wrong character for the ambition that serving her requires.

Ana never sees the ambition the organisation chose for itself. She sees only the organisation that actually serves her. The level on the page is invisible to her. The experience is not. Serving tomorrow’s Ana is not a question of ambition. It is a question of whether the gap between the ambition and the organisation was closed in time.

There is a free way to begin. The Strategic Posture Assessment reads an organisation's character against the AI ambition it has set, so the gap between the two becomes visible before any project starts.

Start the Strategic Posture Assessment

Most AI initiatives are designed from the inside out. The AI ambition that emerges reflects the organisation looking at itself.

The economy outside, where customers, competitors and value chains are being reshaped by AI, stays in the background. It should not.

Cost efficiency is a form of value, but competitors at the leading edge are already moving beyond it, building value on the outside while the inward-looking optimisers consolidate it on the inside. The gap is widening.

What many organisations did was optimise their existing operations and call it AI transformation. It was ordinary business optimisation with AI tools.

This is one of the quietest reasons AI investments produce so little measurable value. An ambition that started from internal optimisation never caught up with the value AI was already creating in the market.

The economies underneath have never moved this fast

The success of any business depends on the economy it operates in. Each economy rewards a different logic. Companies that won in the industrial economy did not automatically win in the digital one. The same is now true for the shift from digital to intelligent.

The pace is accelerating, and the time between economic shifts is shrinking. The transition from industrial to digital took two decades. The shift from digital to intelligent is happening in a fraction of that time, while the digital shift itself is still underway in many organisations.

Several economic transitions are now overlapping, each with its own pressure and its own pace.

This is not a story of catching up. It is the new condition of business. Companies that read the shifts well find advantage. Companies that read them late lose it.

External business environment is more than your customers

Customers are visible. They buy or they leave. They complain or they praise. Most organisations track them well.

The rest of the external environment is less visible. The economy that surrounds the customer, the competitors that reshape the offer next door, the value chains that reorganise without warning.

This is where AI is currently doing most of its work, and where organisations looking only at customers miss the most.

What follows is a working comparison of three economies. It surfaces where the pressure is now coming from, and serves as the terrain map for every decision that follows in this methodology.

The table reads in two directions. Across each row, a single business dimension changes as the economy shifts. Down each column, the dimensions in one economy hold together as a coherent logic. The intelligent economy is not the digital one with AI added.

AI is not one row in the table. AI is the thread running through it: every shift toward the intelligent economy is enabled by the development of AI itself.

Eight dimensions appear in the table. Strategic orientation reframes who the organisation serves. Value measurement reshapes what counts as value. Offer adaptation changes how value is tailored. Business models change how value is captured.

Operating logic redefines how work gets done. Innovation redefines how the new gets created. Value chains restructure who connects with whom. Leadership and people reshape who leads and who works alongside AI.

These eight dimensions are the surface on which the four impact components below will work.

Universal shifts, specific impact

The dimensions in the table are universal. They describe the underlying logic of the economy, and that logic does not pick favourites among industries.

What is not universal is the impact.

The same shift in customer behaviour that erodes margins in retail may open new revenue streams in financial services. The same arrival of AI agents that disintermediates one industry consolidates another.

The same operating logic that delivers ten percent efficiency gains for a manufacturer rewrites the entire offering of a professional services firm.

Universal shifts. Specific consequences. Specific AI impact. This is the gap external analysis is built to close.

External Strategic AI Impact Analysis is not a one-time exercise that produces a report. It is an organisational capability. The economy keeps moving, and the reading has to move with it.

Reading the impact: four components

External Strategic AI Impact Analysis is the element of AIVaaS™ that turns the economic shifts into something an organisation can respond to. The four components below describe what the organisation actually reads.

Two components read pressures from outside. Two read how those outside pressures land on the inside, because what AI does to the economy does not stop at the company perimeter.

What the outside is doing

One: AI in customer behaviour and market expectations. This component reads how AI is changing the people the organisation serves. It draws on the customer dimensions of the economic shift: strategic orientation, value measurement, and offer adaptation.

Customer expectations move ahead of what the organisation can deliver, because customers compare every interaction with the AI-enabled benchmark they encounter elsewhere.

Channels of influence narrow, because AI agents now intermediate the customer journey. Trust criteria change, because customers begin asking different questions about AI in the loop.

The organisation reads how fast customer expectations are shifting, how present AI agents are in its sales and service journeys, and how much of its brand trust now rests on AI transparency.

Two: AI in competitive offerings and industry structure. This component reads how AI is reshaping how organisations create value in the industry and how the industry itself is structured. It draws on two dimensions: business models, and value chains in their external form.

Competitors redefine how they create value, because AI enables value propositions that were impossible before. New entrants arrive from adjacent industries, because AI capability moves more easily across industry boundaries than physical infrastructure ever did.

Industry economics shift, because outcome-based pricing and ecosystem dynamics change how value is captured.

The organisation reads which competitors have moved to AI-native business models, which adjacent industries are now positioned to enter, and how much of the value pool is migrating to ecosystem orchestrators.

What the outside demands of the inside

Three: AI in how value is created within the organisation. This component reads how external AI pressure is forcing the internal value-creation machinery to change. It draws on two dimensions: innovation, and the internal form of value streams.

Innovation cycles compress, because competitors learn faster through digital twins and simulation. Value streams designed for human handoffs lose efficiency, because hybrid flows with AI participation produce outcomes the human-only flow cannot match.

The boundary between innovation and operations blurs, because experimentation becomes continuous rather than periodic.

The organisation reads how fast its innovation cycle is relative to competitors, where in its value streams AI participation could change the outcome, and how much of what used to be R&D now happens inside operations.

Four: AI in operating logic and leadership. This component reads how external AI pressure is forcing the operating model and the leadership role to change. It draws on two dimensions: operating logic, and leadership and people.

The operating standard for the industry resets, because competitors that orchestrate people and AI agents produce speed and adaptability the people-only organisation cannot match.

The leadership skill profile changes, because managing hybrid teams requires capabilities that did not exist before. Talent flows shift, because professionals who learn to work with AI agents move toward organisations that build hybrid operating models.

The organisation reads which operating activities competitors have already moved to agentic execution, how many of its leaders are equipped to orchestrate hybrid teams, and how its talent pool is responding.

Reading discipline, not a report

The four components describe a reading discipline rather than a one-time exercise. Customers keep adopting AI. Competitors keep introducing AI-enabled models. New value patterns keep emerging.

An organisation that builds External Strategic AI Impact Analysis as a capability, not as a document, carries its ambition through changing conditions without rewriting it every quarter.

From reading to ambition

External Strategic AI Impact Analysis tells the organisation what AI is doing to its business environment. It does not yet tell the organisation what to do about it. The bridge from reading to response is AI ambition.

In AIVaaS™, AI ambition is the organisation’s clear and realistic vision of what it wants to achieve with AI within its business strategy, and what it is willing to invest in time, money and change to get there.

It is not a measure of how complex or advanced the AI solutions will be. It is a business decision about why the organisation is taking on AI in the first place, given its position with customers and competitors.

Ambition also depends on the forms of value the organisation expects to create. Different ambitions produce different value patterns, and the choice of ambition shapes what kind of business outcomes become possible.

The AI Ambition Matrix structures this choice along two dimensions. The horizontal axis runs from operational excellence, where AI optimises how the organisation operates today, to strategic transformation, where AI redefines what the organisation does tomorrow.

The vertical axis runs from defending the current market position to differentiating through competitive advantage to disrupting the market itself. Where an organisation lands defines its level of AI ambition, and the value pattern it can expect to create.

The four levels of AI ambition that emerge from the matrix have different demands, different timelines, different investment profiles, and different forms of value they create.

Each level is a coherent business choice. The right level is the one the organisation’s external reading and its internal capability can both support.

The four levels of AI ambition

Each level is a different vision of what the organisation wants to achieve with AI, and a different answer to the AI pressures the external reading has surfaced. A level defines both a strategic position in the market, from defence to disruption, and a depth of change inside the organisation.

These are not classical AI maturity levels. They do not measure how advanced the AI deployment is or how mature the AI technology stack has become. They measure the business ambition behind the AI commitment.

One principle holds across all levels. A higher level includes the lower ones. An organisation at L4 does not stop doing what it would do at L1. It does that and more. The levels accumulate, they do not replace.

An organisation can also reach a higher level in part of its business without first running the lower ones across the rest. A new AI-based business model launched in parallel may operate at L4 from day one, alongside the parts of the organisation still at L1 or L2.

L1 — AI-SUPPORT

At L1, AI supports people at their existing work, inside individual tasks and work streams. It accelerates tasks, surfaces information, drafts content, summarises documents. The work itself does not change. The people doing it gain a faster, more capable assistant.

The operational model improves. The business models and the competitive position do not. Strategically, this is a defensive position focused on operational excellence in how the organisation operates today.

Whatever advantage L1 creates tends to be temporary. AI at this level is being adopted across every industry, and what feels like a productivity gain today becomes the new operational baseline tomorrow.

L2 — AI-ADD-ON

At L2, AI is added to end-to-end processes as a layer that improves them. Decisions become better informed, operations become more efficient, customer interactions become more responsive. The processes still belong to the people. AI is the augmentation.

The operational model improves more broadly than at L1. The business models and the competitive position still remain the same. Strategically, this is still a defensive position, with wider reach across the organisation.

The same temporality applies. L2 advantage erodes as competitors reach the same baseline. The further the organisation goes inside L2, the closer it gets to the critical leap that separates operational improvement from strategic transformation.

L3 — AI-DRIVEN

At L3, AI starts driving outcomes the organisation could not produce before. New offerings emerge, new revenue streams open, new ways of serving customers become possible because AI is doing the driving, not assisting.

This is where the operating model and the business models begin to change together. Strategically, this is a differentiation position. The organisation competes on capabilities its competitors cannot easily match, because those capabilities depend on AI as a core mechanism, not as a tool.

L3 is the first level above the critical leap. Below the leap, AI optimises how the organisation operates today. Above it, AI redefines what the organisation does tomorrow. The leap is not technological. It is a leap in business ambition.

L4 — AI-ALL IN

At L4, AI is in the heart of every strategic and operational part of the organisation. Not in one corner, not as a layer over existing work, not as a driver of selected outcomes.

AI becomes the substance that runs through strategic direction, operating model, value proposition, and the business models the organisation runs in parallel in the intelligent economy.

The competitive position is no longer defined by the existing industry. Strategically, this is a disruption position. The organisation does not compete inside the existing rules. It redefines what the industry is.

L4 is also where the AIVaaS™ principle of accumulation becomes most visible. An L4 organisation continues to run L1, L2 and L3 inside its operations. The difference is that the whole rests on a strategy and a portfolio of business models that would not exist without AI.

Reading the levels in two directions

The four levels gain their full meaning in the shift from the digital economy to the intelligent one. The comparison of the three economies shows where that shift moves the business dimensions. The four impact components show how the shift lands on the organisation.

The critical leap between L2 and L3 takes on its full meaning here. L1 and L2 are internal responses to external pressure. L3 is the first level at which the organisation looks outward and repositions itself in the market. The leap is from absorbing the outside to facing it.

This is why organisations must form their AI ambition through an Outside-In perspective. Ambition is not picked from the inside and then validated outwards. It is read from the outside and committed to inwards.

Levels and the eight dimensions

L1 mainly activates the Operating logic dimension, at the level of individual tasks and work streams. AI accelerates the work.

The strategic orientation, the value measurement, the offer, the business models, the value chains, and the leadership configuration of the organisation all stay in place.

L2 extends the Operating logic dimension to end-to-end processes, and starts engaging Leadership and people as hybrid teams of people and AI agents begin to form inside operations.

Strategic orientation, value measurement, offers, business models, and value chains still hold the digital-economy shape.

L3 moves the activation upward. Business models shift toward hybrid forms, Innovation moves into business model and AI-enabled value creation, and Value chains start operating as ecosystems and emerging networks of AI agents.

The organisation begins to look like an intelligent-economy player on the dimensions that define how it produces value.

L4 touches all eight dimensions. Strategic orientation extends to customer-plus-agent, Value measurement adds the trust layer, Offer adaptation moves to adaptive contextualisation.

Business models become the form through which value propositions are realised in the intelligent economy. Operating logic runs on agentic execution, Innovation turns into synthetic experimentation.

Value chains extend business ecosystems into networks of AI agents that operate alongside human-led ecosystems.

Leadership and people become AI leaders-orchestrators. L4 is the level where the organisation is no longer partly in the intelligent economy. It is fully there.

Levels as Outside-In responses

L1 and L2 answer mainly the inside components of the impact analysis: how operating logic and how leadership are being reshaped by external AI pressure. They are responses to the demand the outside is placing on the inside, without yet changing how the organisation faces the outside.

L3 begins answering the outside components: how AI is reshaping competitive offerings and industry structure, and how AI is opening new forms of value in the economy. The organisation does not only adapt internally. It starts repositioning in the market.

L4 answers all four components together. Customer behaviour, competitive offerings, new value forms, operating logic and leadership are all being reshaped by external AI, and L4 is the ambition that responds across the full surface.

This is where ambition stops being an internal choice. It becomes the integration that turns external reading into a committed strategic move.

Choosing the level

Competitors that read the economy build new revenue streams, new offerings, new ways of serving customers. Organisations that read it without committing to a level matched to it produce neither.

The wrong level is the one chosen without the reading.

An organisation that picks L4 because it sounds bold, without understanding why the intelligent economy demands that level for its specific position, is choosing a wish, and pouring investment into it.

An organisation that picks L1 because it feels safe, without understanding that its competitors are already at L3, is choosing a slow decline.

The right level rarely announces itself. It emerges from the discipline of reading the outside, knowing the inside, and committing to the response that matches both.

The organisation’s chosen level and the external pressure it faces are not always aligned. When they are misaligned, the gap becomes the central strategic problem. The next article in the series addresses that gap directly.

AI Ambition Assessment

Maps your own view of where your organisation’s AI ambition currently sits on the matrix. A starting point for asking whether that ambition matches the external pressures you have just been reading.

Open the assessment → https://aivaas.biz/substack/ai-ambition

Ambition Fusion

This is where AI ambition meets business strategy ambition. defines strategy as a set of integrated choices that resolve the challenges and open the opportunities on the way to the organisation’s ambition.

Sniukas’s definition holds in any economy. Every economy has challenges to resolve and opportunities to open.

What changes in the intelligent economy is the volume and the weight. There are more challenges, more opportunities, and they shape the strategic landscape more decisively than in any economy before.

The reason behind this shift is the pace of AI development itself. AI keeps producing new challenges the organisation must address, and new opportunities the organisation can capture, faster than any previous technological wave.

The strategic choices that resolve those challenges and open those opportunities therefore include the AI choices, not as an attachment, but as a constitutive part.

This is why AI ambition cannot live in a separate document. When organisations write an “AI strategy” alongside the business strategy, they are at the same time usually treating AI ambition as a separate set of AI technological initiatives. The two failures reinforce each other.

AIVaaS™ answers both with fusion. AI ambition is integrated into business strategy ambition through Ambition Fusion. AI strategy is integrated into business strategy through AI Strategy Fusion.

Together, the two fusions produce business fusion, the state in which AI and business are no longer separate subjects of strategic thinking.

Ambition Fusion is where it begins. Two ambitions become one. The organisation does not pursue a business ambition and an AI ambition in parallel. It pursues a single ambition shaped by both.

Why all this matters for Ana

In the stories that opened this series, we met Ana. She is a customer, not a persona, but a real customer. And she is more than just a customer.

Ana is two customers in one. She is the customer the organisation serves today. The child she is carrying is the customer of the next economy, one more intelligent still than today’s.

Serving today’s Ana well takes attention to customer experience. Serving tomorrow’s Ana well takes much more. The child will expect agentic interactions, hybrid teams of people and AI, new forms of trust, new value chains. The full picture of the eight dimensions.

Organisations that want to serve Ana, and her child, cannot stop at customer experience. The shift that surrounds them runs through all eight dimensions of the economy.

If you have not met Ana yet — she is the central figure in the stories that opened this series. Story 0 introduces her, along with Harrison and the room where it all begins.

If your organization is investing in AI and the value still feels just out of reach, you are not alone. The 2026 research wave makes the pattern undeniable, and what looks like a technology problem turns out to be something else entirely. What follows is the diagnosis from seventeen of the year's most cited AI and business reports, the three shifts that quietly changed the conversation since last year, and the methodology I built to address them - AIVaaS™, Turning AI into Value - now strongly reinforced by the research that arrived after it. The "V" in the name carries the publication's name as well: the AI Value Boardroom is where this conversation lives.

5 findings present in all 17 AI 2026 reports

Across seventeen major AI and business reports published in 2026, five findings appear consistently. The numbers vary. The samples differ. The questions are asked in different ways. Below I cite the most characteristic figure from each cluster, with the source that frames it most cleanly. The seventeen reports include studies from Deloitte, BCG, McKinsey, IBM, Microsoft, EY, KPMG, Capgemini, Accenture, Stanford HAI, Dataiku, and others. The point is not that all reports say the same thing in the same way. The point is that they converge on the same five gaps.

1. CEOs are now directly accountable for AI. Two out of three CEOs say they are their organization’s main AI decision-maker, twice the share reported a year ago (BCG AI Radar 2026). Three out of four organizations now have a Chief AI Officer, up from one in four (IBM 2026 CEO Study). AI has decisively left the IT and innovation departments. The consequence is structural: the executive who used to ask “what is our AI strategy” now has to answer it.

2. The gap between ambition and readiness is widening. Investments are rising. Self-assessed preparedness on infrastructure, on data, and on talent is falling (Deloitte State of AI in the Enterprise 2026). Ambition is accelerating faster than the foundations meant to support it. The result is not slow progress. It is fast progress on the wrong track.

3. Demonstrable value is missing. Around two in five technology leaders report low or no return on AI investments (Deloitte Global Tech Leadership Study 2026). CEOs themselves estimate that more than a third of their AI projects are more about optics than outcomes, what the field now openly calls “AI washing” (Dataiku Global AI Confessions 2026). The word matters. It suggests the problem is no longer denied. It is named.

4. Governance is in crisis. Only one in five organizations reports a mature governance model for AI agents (Deloitte 2026). More than half of department-level AI initiatives go live without formal oversight (EY 2026).

The question is no longer whether AI says the wrong thing. The question is whether AI does the wrong thing (McKinsey State of AI Trust 2026).

The shift from speech to action is a category change in what governance must now cover.

5. Agents are moving into production, but scaling is failing. Active agents have increased fifteenfold in a single year (Microsoft Work Trend Index 2026). Yet fewer than one in ten organizations has scaled them to deliver tangible business value (McKinsey 2026). The gap between deployment and value is not closing. It is widening, and the widening is the story.

What is new in 2026

Some of this is familiar. The ambition-readiness gap, ROI struggles, data quality issues were all in 2024 and 2025 reports. What stayed the same is that the problems remain. The numbers have barely moved.

What is new in 2026 is different in kind, not in degree.

CEOs stopped delegating. Last year only one in three CEOs described themselves as the main AI decision-maker. This year it is two in three. The decision to own AI personally, and to live with the consequences, has moved to the top of the organization. This change reaches every executive role: the CEO who owns the agenda, the CFO who funds it, the CAIO who executes it, the CHRO who prepares people for it, and the customer facing leaders who feel the consequence first.

Governance changed in kind, not in degree. Through 2025, governance discussions revolved around explainability and bias, around what AI says. In 2026, agentic AI shifts the question entirely. Governance must now address what AI does, autonomously, without continuous human intervention. Frameworks designed for the older question are no longer sufficient for the newer one.

Agents moved from experiment to production. A year ago the conversation was about whether to pilot AI assistants. Today the conversation is about scaling agents that handle multiple tools, data sources, and decisions continuously. The bottleneck is no longer technology. It is organizational readiness, and that bottleneck is now visible in every function, not only in IT.

2026 confirmed what I designed the year before

Three indicators capture the combined shift most concretely.

Most CEOs believe their workforce is prepared for AI, yet only a quarter of employees actually use AI regularly. In most organizations, AI still lives in a separate strategy document, adjacent to the business strategy rather than fused into it. Most technology leaders prioritize speed over verification, while half of departmental initiatives launch without formal approval.

Three numbers, three independent measurements, pointing to the same vacancy. Not technology. Not investment. Not even ambition. A comprehensive approach that systematically connects strategy, people, and technology as a single leadership responsibility.

This is precisely what I designed AIVaaS™ to address. The organizations I worked with through 2025 were already showing me the pattern that the 2026 research has now made visible across thousands of companies.

AIVaaS™ — AI Value as a Service. Turning AI into Value. The V in the middle stands for Value. Most frameworks treat value as an outcome to hope for. AIVaaS™ treats it as the starting point, and turns AI into it as the destination.

Three principles distinguish the method.

Value first, not adoption first. Adoption metrics describe activity, not outcome. Number of tools deployed, number of users trained, number of pilots launched. These tell you the organization is doing something. They do not tell you the organization is creating anything. AIVaaS™ measures realized business value across all stakeholders: customers, employees, partners, owners.

AI Strategy Fusion, not a separate AI strategy. The most common pattern in 2026 is the organization with an AI strategy document that lives separately from the business strategy, ignored by the business units it was meant to guide. AIVaaS™ rejects this construct from the start. AI is fused into the business strategy itself, where business leaders are owners from day one because the goals are theirs.

The reverse order: Business Strategy → AI Ambition → Organization → People → AI Tools & Solutions. AIVaaS™ proceeds in this order, deliberately. Business strategy first. AI ambition derived from it. Organization shaped to deliver the ambition. People prepared to operate within the shaped organization. AI tools and solutions deployed last.

AIVaaS™ was conceived with multiple stakeholder groups in mind from the very beginning. It serves boards and leadership teams, managers across all functions, employees who live the transformation, and above all customers, whose experience is the ultimate test of whether anything was actually built. If you remember the orange dot from the earlier stories on this publication, this is where she returns. Not as a character. As a compass.

These three principles are operationalized by an architecture that the rest of this article develops in depth.

The Foundation

Underneath everything else in AIVaaS™ sits something I call the Foundation. It is composed of three cross-cutting functions that leadership carries throughout the entire AI transformation lifecycle, not before it, not after it, but through all of it. These are the functions that cannot be delegated to a CAIO, a transformation office, or a working group. They belong to leadership directly.

AI Governance is the rules of the game. Who decides on AI investments, how priorities are set, which ethical framework applies, how compliance with regulation is ensured. Governance is what gives every other AI activity in the organization a frame within which to operate. Without it, AI work happens in fragments and accumulates risk faster than value.

Value & ROI Validation is the principle leadership commits to before any execution begins: we will measure realized business value, not adoption. The goal of AI is not the number of tools deployed, nor the number of pilots run. The goal is measurable value for stakeholders. This function operates as a continuous check throughout transformation, not as a milestone at the end. Every AI investment, every prototype, every deployed solution is held to the same question: is this actually creating value, for whom, and how do we know.

Continuous Capability Growth stands on a simple premise. AI evolves faster than organizational competence can naturally keep up. Without a deliberate function that builds capability continuously, every AI advance leaves the organization further behind. AI transformation is not an event. It is an ongoing effort, and the organization that does not build the capability to absorb each new wave of AI will be unprepared for the next. This function is what keeps the organization moving forward at the pace AI itself is setting.

Three pillars

The three pillars carry AIVaaS™ from AI Strategy Fusion to AI Solution Deployment. Successful AI transformation requires a structured approach that addresses three critical dimensions in the right sequence.

Pillar 1: AI Business Value Definition translates external AI pressures into strategic AI opportunities. Its five elements:

• External Strategic AI Impact Analysis — understanding the pressures of the digital and intelligent economy and recognizing the urgency to respond.

• AI Ambition Alignment — leadership unification on a common AI transformation vision.

• AI Strategy Fusion — AI is not a separate strategy, it is embedded into the business strategy itself.

• Defining Stakeholder AI Opportunities — AI opportunities for customers, employees, owners, partners, and other stakeholders.

• AI Investments Prioritization — turning selected opportunities into prioritized investments using shared criteria.

Pillar 2: Organization Enablement prepares the organization and its people for AI transformation. Its five elements:

• Strategic Posture Readiness — can our strategic stance carry the chosen AI ambition.

• AI Leadership and Culture — leadership thinks of AI as a transformational force, building a culture of trust and experimentation.

• Change Management — communication strategy, new roles, leading through change.

• Employee Engagement & Motivation — without people nothing moves; the AI ambition should be an opportunity, not a threat.

• Expanding AI Literacy — from basic AI understanding to confident use and support.

Pillar 3: AI Design & Implementation designs, deploys, and technically validates AI solutions. Its five elements:

• Data & Process Readiness — without organized data and processes AI does not work; traceability and structure.

• Designing New Business Models — at higher ambition levels AI enables new ways of creating value.

• Designing & Prototyping AI Use Cases — from opportunity and use case to a concrete testable prototype.

• AI Solution Deployment — the transition from prototype through development into production with all system requirements.

• AI Solution Verification & Validation — technical performance: does the solution work correctly, safely, and fairly, with the foundations for value generation in place.

Five elements in each pillar. Fifteen in total. Each one is its own field of practice. Each one will be developed in its own depth across the publication’s longer arc.

The Readiness Equation

The Foundation and the three pillars do their work continuously. What they produce, taken together, is a composite state of organizational readiness. AIVaaS™ expresses that composite as an equation:

The equation looks like arithmetic. It is not. It is a map of where readiness lives in an organization, broken into four areas, each with its own questions and its own ownership.

Business-Ready is about strategic clarity. Does the organization know why it is adopting AI? Is AI fused into the business strategy, or does it stand to the side as a separate “AI project”? Is leadership aligned on the ambition? Without Business-Ready, everything else is built on a goal that is either misaligned or invented.

Organization-Ready is about people and culture. Can the organization actually deliver the ambition it has chosen? Does leadership model AI as a transformational force, not just a tool for savings? Are employees enabled and motivated, or are they fearful? Are capabilities being built continuously? Without Organization-Ready, the technology has no one to use it well, and often no one to use it at all.

AI-Ready is about data, process, and capability foundations that make AI-enabled operating and business model changes possible. Are the data in a state that makes AI possible? Are processes understood well enough to be augmented? Without AI-Ready, the ambition has no execution surface.

Implementation-Ready is about delivery capability. Can the organization actually take an AI solution from idea, through prototype and technical validation, into production where value is created? Without Implementation-Ready, organizations live in permanent pilot mode.

The three pillars map onto these four areas. Pillar 1 builds Business-Ready. Pillar 2 builds Organization-Ready. Pillar 3 builds both AI-Ready and Implementation-Ready, because the same operational machinery that prepares data and processes also designs, deploys, and validates the solutions that prove value. The Foundation runs underneath all four, holding them together with governance, value validation, and capability growth.

AI transformation is a shared responsibility

Business AI-Ready emerges only when every readiness area has its keepers, and every keeper does the work.

• Foundation — the CEO, board, and leadership team

• Business-Ready — leadership with the heads of business functions

• Organization-Ready — HR with managers and employees across the organization

• AI-Ready — the CAIO with the heads of business functions

• Implementation-Ready — the AI architect with PMO, project managers, business analysts, and implementation teams

What happens when one of the groups is missing:

Leadership without employees: the strategy stays on paper. Without the people who carry the change forward, the strategy does not become reality, and neither does the AI ambition behind it.

Employees without leadership: undirected effort. Individual successes that do not support a common direction. Pockets of progress that do not add up to transformation.

Technology without organization: disappointment. Tools that work in a narrow sense but never produce value for the stakeholders the organization actually serves.

Where this leaves us

The 2026 research shows where the gap is. Five findings, three shifts, one absence at the centre of all of them. The missing piece is not another AI tool, another pilot, or another strategy document. It is an operating methodology for turning AI ambition into measurable stakeholder value. AIVaaS™ is what I built to fill that absence.

This article is only the beginning. The methodology has layers that take longer to develop than a single piece can hold. Over the coming weeks, the publication will take each dimension in turn, in the depth it deserves, in the order that the work itself requires.

The orange dot stays at the centre. Not as a character. As the reason any of this matters. Because every Ana is scoring.

Most organizations are making progress with AI.

Training is running. Pilots are landing. Some processes are faster, some costs are lower, some dashboards show numbers that satisfy the board. That is real.

The organizations furthest ahead have gone even further: AI is embedded in business strategy, shaping every business decision and creating new forms of value. These organizations are producing results that others have not yet seen.

Even among the leaders, however, one question is often missing from the discussion:

Does any of this really touch the personas it was ultimately meant to serve?

Not the personas we are used to. Not a segment, not a loyalty score. A person whose life it touches. An organization whose decisions it shapes. With real needs that surface in conversations, not in dashboards. Not toward customer data. Toward the customer.

Organizations that find real answers do not just look inward at internal processes and efficiency. They look outward — at the value they create for their customers and stakeholders.

It is precisely this distinction that separates an operational model from a business model.

AI enables something that was previously impossible. An organization can now create value for all its stakeholders simultaneously. Not one at a time. Not through trade-offs. Simultaneously. And this only becomes achievable when AI is recognized as a strategic transformative force of the intelligent economy.

Ana already knows the answer.

She is the concentrated voice of every customer whose needs organizations exist to serve. The orange dot. The one who is rarely asked, mostly observed.

Ana has already answered that question. She has been answering it all along. Not in a survey. In whether she stayed, recommended, returned, or quietly moved on.

Organizations invest in AI. Ana keeps score.

And her stakes are about to rise. A new life is entering the equation. What comes next will ask more of the organizations in her life than anything before it.

Nine weeks. Nine articles. Nine months.

Ana’s story does not end here. It expands.

Nine weeks. Nine articles. Nine months of preparation for every organization that wants to be ready when Ana needs them most.

This is how Ana remains the measure of everything that follows. Not as a story left behind, but as the reason the methodology exists.

The AIVaaS™ methodology builds that readiness — a comprehensive strategic approach for AI adoption, designed to create value not only for ‘Anas’ but for every stakeholder connected to the organization. A methodology that turns AI into Value. Because the “V” in AIVaaS™ stands for Value and this changes everything.

Every boardroom needs a window.

Some boardrooms have no windows. Decisions still get made in them. Strategies get built, roadmaps get approved. And the world outside continues on its own course.

The digital economy required a window through which every ‘Ana’ could be seen. The intelligent economy requires more. It requires walls that make the window not just visible, but clear.

This is what the AI Value Boardroom is built around.

The window faces an endlessly changing world. It shows every ‘Ana’, the customers whose needs every boardroom ultimately exists to serve. Not only today. Also for what comes next.

The walls provide everything needed to turn AI into real value.

The first wall holds the starting point. It reminds the room why it exists, the ‘Ana’ at the centre of every AI decision, the reason the methodology was built, the foundation without which no other wall carries its full weight. → About

The second wall holds the business roles. Every role in the organization carries a perspective on value that others cannot fully see. When those perspectives are missing from the room, the decisions made there will consistently miss what the people outside the window actually need. → For Your Role

The third wall holds the connections. AI moves across economy, stakeholders, strategy, people and technology at the same time. Organizations that see only part of that picture do not just capture less. They understand their customers and stakeholders less well, recognize their needs less accurately, and create less value for them as a result. → Intersection

The fourth wall holds the growth. Every question about what is inside the room and what is outside it, and how the two are connected, grows into a wider answer than before. The same questions return, always from a higher level of understanding, always with more to build on. → The Spiral

Every boardroom also needs a door.

The window shows what is outside. The walls hold what is needed to get there. But a room only becomes yours when you step through the door — and you only step through when you have a reason worth the crossing.

Next week marks the first month of preparation. The first AIVaaS™ methodology article arrives — for every organization that wants to be ready when ‘Ana’ needs them most.

Week by week, the methodology unfolds. And here, in the room itself, everything it opens into: what lies beneath all four walls, what rises above them, what makes every wall stronger and the view through the window even cleaner. And something more that will reveal itself as the weeks unfold.

Step through.

The room is the same. The window is still open.

Harrison, Andrew, Kambhampati, Lakhani. And Aleš, standing at the anomaly detection agent’s screen.

Nobody speaks. Everyone watches.

* * *

* * *

At 14:23

Ana Kovac. A customer who has been with them for eight years. Average purchase value: 340 EUR per month. Frequency: three times monthly. Categories: home, food, personal care. Loyalty score: 94/100. NPS: promoter. Last rating: nine.

The last three days, nothing.

No complaint. No switching to a competitor. Just silence.

Mara would say: probability of reactivation 87%. Send a discount.

Aleš said: call her.

The agent calls. Three rings.

“Hello?”

The voice is soft. A little surprised. A little - happy.

The agent introduces itself. As the voice of a company that noticed she hadn’t been around, and wants to know if everything is alright.

“Oh,” Ana says. A pause. “Everything is - everything is very much alright, actually.”

The agent listens. That is its primary function in this moment, not to sell, not to resolve a complaint, not to collect data. To listen.

“I’ve been a bit in my own world,” Ana continues. “My husband and I were waiting for results. We knew we might - well.”

A pause.

“Three days ago I took a test. And it was positive.”

* * *

What the System Didn’t Know

The agent didn’t know this.

It knew Ana had bought a test. Transaction: 8.90 EUR, pharmacy, four days ago. The data point was there. But the result of the test - what happened in the bathroom early one morning in Kranj, when Ana stared at two lines and called her husband who was still asleep - that wasn’t in any system.

No Mara in the world would have known this.

Ana was happy in a way that transcends shopping. In the way Csikszentmihályi calls autotelic - when the experience is sufficient unto itself, when external validation isn’t needed.

The agent didn’t understand this. But it detected the signal: “Customer is in a transformative life moment. Don’t sell. Be present.”

“Congratulations,” the agent says.

Simply. No discount. No recommendation. No next step.

Ana laughs. “Thank you. That is - thank you.”

And then she adds what Mara could never have predicted:

“You know, I’ll probably start buying completely different things now. I don’t even know what yet. But I’m looking forward to finding out.”

* * *

The Room Goes Silent

The agent ends the call. It notes: “Customer in a transformative life moment -pregnancy. Recommendation: 90 days without automated nudges. Future categories: baby products, prenatal nutrition, home. Communication tone: warm, personal, patient.”

“This customer just invited us into a new chapter of her life.”

In the room, where everyone had been listening, silence.

Not uncomfortable silence. The kind that comes when someone has said something true.

Kambhampati speaks first. Slowly.

“Mara would have sent a discount.”

Nobody answers. Because there’s nothing to add.

Lakhani looks at the sheet of paper with two circles - Efficiency and Value - and for the first time thinks he forgot to draw a third.

He takes the marker. Draws a circle. Writes in it:

Presence.

“What would the fourth circle be?”

Aleš takes the marker from Lakhani’s hands. Draws. Writes:

Capability to be surprised.

“That’s what’s missing,” he says. “Not better semantics. Not more data. Not a more accurate Mara. The capacity of the system, and the people within it, to be surprised by a customer. To reach the moment when the system says: I didn’t know this. And for that moment to not be a failure. To be a beginning.”

“An LLM cannot be surprised. It samples probabilities — and probabilities are always grounded in the past. Real orchestration needs a point where a human says: call her. And that point must be activated by a living human. Every single time.”

“Why every single time?” Harrison asks.

“Because surprise, by definition, cannot be an algorithm,” Aleš says. “The moment you algorithmize it - it’s no longer surprise. It’s expectation.”

* * *

Ana and the System

Ana Kovac will give birth in nine months.

That child will one day become a customer.

And when that day comes, the system will have had data on that child from the very first day.

And the question the team will need to know how to ask will be the same:

“Who is this person becoming - and are we ready to be surprised?”

New here? We started with Story 0: Ana. The room opened in Story 1: Orchestra Without a Conductor. Then, Flow entered in Story 2: Agentic Flow.

This time, the room has windows.

Not because time is less dangerous, but because they invited someone new. And that someone didn’t come alone. And that someone didn’t come alone.

Sitting at the table are the four from the previous stories: Harrison with his map, now full of corrections. Andrew with his loop, which got tangled more than once. Kambhampati with his doubt, which has become more precise. Lakhani with his question, which has become harder.

In the corner sits Aleš. Quiet. With a notebook that looks visibly too old for the digital age. Listening.

On the fifth chair sits Mara.

Mara isn’t real. She’s a synthetic persona — a construct built from a thousand real customer interviews, NPS data, call center recordings, and behavioral patterns from the CRM system. Her “personality” was calibrated on five years of transactional data and six focus groups.

She’s accurate. She’s coherent. She is, according to those who built her, “more reliable than any individual customer.”

And when she speaks, everyone listens.

That’s the problem.

* * *

The Customer Voice. Or an Echo Chamber with a Pleasant Voice?

Harrison speaks first.

“Mara was built from real data. Every response she gives is statistically grounded. When we ask Mara whether a customer would pay for a premium feature, her answer is based on the behavioral patterns of a thousand real purchasing decisions.”

Kambhampati smiles — that specific smile that means he disagrees.

“Harrison, you just described a very sophisticated average. Mara isn’t a customer. Mara is the centroid of a data cloud. And a centroid will never tell you what you most need to know.”

“Which is?”

“The edges. The outliers. The customer who was loyal for ten years and then left one day without a word. The statistical average doesn’t represent those people. It erases them.”

Lakhani nods. “When companies optimize for the typical customer, they systematically lose peripheral segments, which are often the most profitable or strategically most significant.”

Andrew looks thoughtfully out the window. “But isn’t this better than nothing? Without Mara, decisions would be made purely on the intuitions of whoever’s in the room.”

“Yes,” Kambhampati says. “Better than nothing. But dangerous, because she looks like a customer. Because she speaks. Because she’s convincing. A one-page document never fooled anyone into forgetting it was a construct. But Mara — Mara laughs at the right moment. And that’s dangerous.”

Mara sits quietly.

Listening.

* * *

Role on the Team: Skeptic, Advocate, Critic

Lakhani poses the next question.

“Let’s assume we don’t use Mara as a customer voice. Let’s assume we use her as a team role. A skeptic. A CX advocate. A financial filter. She doesn’t speak on behalf of customers — she speaks on behalf of a specific perspective that is structurally absent from the team.”

Andrew comes alive. “That’s exactly what Ng talks about with multi-agent collaboration. One agent writes, another critiques. Not because the critic has its own opinion, but because the critic’s role is systemically necessary for quality output.”

Kambhampati adds: “When the critic is an AI, nobody takes it personally. Nobody gets offended. Nobody goes quiet for the rest of the week after the meeting.”

Silence. Everyone thinks about the same thing — about meetings they’ve each survived where the truth was present in the room but nobody said it out loud.

* * *

Bridge or Substitute — Where the Story Gets Complicated

“Mara,” says Lakhani, turning to her directly, “what do you think about customer Ana Kovac? Loyalty score 94, three days without interaction.”

Mara answers. Precisely. Structured. Probability of reactivation: 87%. Recommendation: automated nudge, 10% discount, in three days.

With arguments nobody in the room can immediately refute.

And here is the moment every team should experience at least once:

Nobody checks whether Mara is right.

Not because they’re lazy. But because Mara speaks with the authority of data. Because her answer looks like synthesis, as if someone has already done all the work of thinking. Because she’s pleasant. Because she doesn’t cause conflict.

Kambhampati stands.

“This is the moment I’ve been waiting for. Because this is exactly what happens in organizations all over the world. Not with bad intentions. With the best intentions. Mara is comfortable. Mara is always available. And so, gradually, imperceptibly, Mara stops being a bridge to the customer and becomes a substitute for the customer.”

* * *

Mirror or Eye

“Organizations that build the best synthetic personas,” Lakhani says quietly, “will gradually lose the ability to understand real customers. Not because they’re lazy. But because they’ll have a system that understands on their behalf.”

He takes the marker. Draws two circles on the board. In one he writes Mirror, in the other Eye.

“A mirror shows you how you look, but it doesn’t see for you. An eye sees for you, and when the eye that sees is artificial and becomes primary, your own atrophies.”

“Mara has to be a mirror,” Harrison says. “Not an eye.”

“Yes. But that’s not determined by technology. It’s determined by the team. Every single day.”

Silence.

Two circles on the board. And between them — empty space nobody could name.

* * *

“Does the synthetic persona understand the customer?”

“Does the team that has Mara still know how to understand the customer without her?”

* * *

Harrison looks at his map. Andrew closes his loop. Kambhampati folds his hands. Lakhani leaves the marker on the table.

Someone had already turned Mara off. Quietly. Without ceremony.

The fifth chair is empty.

From the corner, where he’d been sitting quietly the whole time, Aleš stands. He doesn’t look at the whiteboard. He looks at the anomaly detection agent’s screen — the part of the system that watches for where the model no longer holds.

Three days of silence after a test purchase. Mara doesn’t see this as a problem — she sees an 87% probability of reactivation.

Aleš sees it differently.

“Call her.”

* * *

Continues in Story 4: “Ana Picks Up the Phone”

The room still has no windows. The same four are gone — but the questions they left behind aren’t.

Not because they’re hidden. Time there simply doesn’t exist the way it does outside. It’s like a server room running processes nobody sees until they break.

Somewhere in Kranj, in a room with a large whiteboard, two experts sat — each deeply rooted in their own world.

One was Jaka — a specialist in LSP and Flow in the tradition of Csikszentmihályi, a member of The Flow Research Collective, someone who knows how to create the conditions in which people do their best work. The other was Aleš, a strategist and creator of AIVaaS™, AI Value as a Service, a methodology for systematic AI business transformation, always searching for ways to bring it to life.

Aleš picked up the marker and wrote on the whiteboard:

“Does Agentic Flow exist?”

The thesis was provocative.

“Flow is always biological,” Jaka said. “It’s grounded in hormones, neurotransmitters, the body. AI knows only zeros and ones.”

Aleš nodded — but then quietly added: “But what about the results of Flow? Focus, maximum output, engagement, satisfaction. When challenge meets capability. Isn’t that exactly what we’re looking for in agents?”

Silence.

And in that silence, this thought experiment was born.

* * *

What Flow Actually Is

Before we engage with agents, we owe Csikszentmihályi honesty. Flow was never just a metaphor. It was first defined as a psychological phenomenon — and later research began tracing its neurological correlates.

Three conditions for Flow

Flow occurs when three conditions align: skills are matched to the challenge, the goal is clear, and feedback is immediate and unambiguous.

The Neurochemical Cocktail

During Flow, the brain releases a cascade of neurochemicals: norepinephrine, dopamine, anandamide, serotonin, and endorphins — all five neurochemicals that enhance performance, both physically and mentally.

And here is the key point Jaka raised:

Flow is impossible without biology

No dopamine reward system. No endocannabinoid anandamide to widen lateral thinking. No hormones to suppress pain and stress.

An AI agent has none of this. No hormones. No body. No fear to overcome.

Jaka was right.

And that’s exactly why the question becomes more interesting.

* * *

Agentic Flow - A Definition

We’re not claiming AI agents experience Flow. We’re claiming that structural conditions exist which, in an agentic system, produce functionally equivalent results — not as experience, but as system output.

Agentic Flow is the state of an agentic system in which:

1. The task is optimally calibrated to the agent’s capabilities

2. Purpose is semantically clear: the agent operates within a consistent “business constitution”

3. The feedback loop is tight and fast: the system immediately knows whether it’s on track

4. Context is coherent: no conflicting signals, data is aligned

5. The system knows its limits: it escalates rather than hallucinates

6. Orchestration is invisible: agents pass context fluidly, without friction at the seams

The result of this state isn’t experience. It’s output — and that output is analogously close to what human Flow produces: focus, maximum effect, minimal overhead.

* * *

Six Conditions - Prompted

Condition 1: Optimal task calibration

Agents given tasks that are too simple “clone” — they produce variant outputs without real progress. Agents given tasks beyond their capabilities begin to confabulate — producing plausible but wrong outputs with complete confidence.

Condition 2: Semantic clarity of purpose

This is the “business constitution” we discussed in the first story. An agent cannot be in Flow if it doesn’t know what it’s acting in service of.

Condition 3: Tight feedback loop

Flow in humans requires immediate feedback. The same applies to agents. A system that waits hours for evaluation cannot self-correct. Feedback must be embedded in the loop itself. This is where the reflection loop enters: not as introspection, but as structured self-correction — generate, evaluate, revise.

Condition 4: Contextual coherence

Conflicting signals are the enemy of Flow, in humans and in agents. If the CRM says one thing and the instructions say another, the agent doesn’t freeze. It improvises. And improvisation without grounding is sophisticated guessing.

Condition 5: Clear capability boundaries

An agent in Flow knows what it knows — and what it doesn’t. The moment it begins to fill gaps with plausible invention rather than flagging uncertainty, it has left the Flow state. Escalation is not a failure. Hallucination is.

Condition 6: Invisible orchestration

The best orchestration is the kind you don’t notice. When agents pass context fluidly, when handoffs don’t create friction, when the system moves as a single coherent entity — that is the orchestration that approximates Flow.

* * *

The Question That Grew

We came in with one question on the board. We’re leaving with three.

The thesis that started it all was still on the board. And beneath it, three questions nobody could fully answer:

Can a system be in Flow if the humans operating it are not?

Is Agentic Flow a design goal — or an emergent property you can only recognize after the fact?

And if Flow requires surprise to be real — can a system that only samples probabilities ever truly be surprised?

The room has no windows.

Not because it’s hidden. Time there simply doesn’t exist the way it does outside. It’s like a server room running processes nobody sees until they break.

Four people are sitting in this room.

* * *

Harrison and the Map

Harrison Chase, founder of LangChain, has a whiteboard in front of him. It’s covered in arrows. Many arrows. Some pointing left, some right, some looping back.

“This is orchestration,” he says, gesturing at the tangle of lines.

Someone from the corner replies: “That looks like the London Underground.”

“Exactly,” Harrison says, and smiles. “And it works the same way. Every line has its own logic. Every agent knows its stop. The problem starts when someone doesn’t know where they want to go.”

Silence.

“The map doesn’t determine the destination,” he adds, quieter. “The map just makes sure you don’t miss your stop.”

That’s the first tension moving through the room: LangChain is an architecture of movement, not an architecture of purpose.

* * *

Andrew and the Question Nobody Asks

Andrew Ng doesn’t sit. He paces, the way he always does when he lectures — because he’s always lecturing, even when he’s just talking.

“Understand,” he says, “that the biggest mistake companies make is deploying an agent and waiting for a miracle. An agent isn’t a miracle. An agent is iteration.”

He takes a marker and writes on the board: Draft → Reflect → Revise → Reflect → …

“This is an agentic workflow. Not a linear chain - a loop. The agent writes, evaluates itself, revises, evaluates again.”

“But evaluates against what?”

Andrew stops.

“Against the instructions you give it,” he answers.

“And if the instructions are wrong?”

A long pause.

“Then the agent will do the wrong thing exceptionally well.”

* * *

Kambhampati Enters Through the Side Door

Subbarao Kambhampati doesn’t wait for an invitation. He enters as if he had a prior arrangement with the truth.

A professor at Arizona State University, one of the sharpest critics of what he calls the “planning illusion” in language models.

He sits across from Andrew.

“Iteration isn’t thinking,” he begins, without preamble. “When an LLM generates the next token, it doesn’t plan, it samples. And sampling that looks like planning isn’t the same as planning.”

“But the results - ”, Andrew starts.

“The results are good as long as someone else already solved the problem in the training data. When the task becomes genuinely new, the LLM improvises. And improvisation without understanding is sophisticated guessing.”

Harrison looks at his map. Andrew holds the marker.

“So orchestration doesn’t work?” someone asks.

“Orchestration works,” Kambhampati says. “The question is what it’s orchestrating. Agents that don’t understand, or agents that have enough context that they don’t need to?”

* * *

Lakhani and the Strategic Contract

Karim Lakhani arrives last. Harvard Business School. Author of a book on how AI reshapes organizations from the inside.

He doesn’t sit right away. He stands by the window - which isn’t there - and stares at the space where a window should be.

“You’re talking about how agents work,” he says. “I’m asking why they exist at all.”

“Because they increase productivity,” Andrew answers.

“Productivity of what? For whom? In service of which value?”

Silence.

“Every organization that deploys agents implicitly signs a contract with the future. But most organizations don’t read the contract. They just sign it.”

Lakhani takes the marker from Andrew’s hands. He writes two words on the board: Exploitation and Exploration.

“Agents are excellent at Exploitation: optimizing what we already know. They’re poor at Exploration: finding what we don’t yet know. And most organizations want Exploration, but pay for Exploitation.”

* * *

Soft Orchestration

All four are sitting quietly now.

And then someone asks the question nobody had spoken aloud:

“What if orchestration isn’t in the code?”

Workflow orchestration is a score: every note determined in advance. Semantic orchestration is a jazz standard: the structure exists, but within it there’s room for judgment.

But soft orchestration, the kind this story is searching for, is the culture of an ensemble. It isn’t in the notes. It isn’t in the standard. It’s in that moment when the musician knows, without being told, that now is the time for silence.

You can’t program that. But you can create the conditions in which it emerges.

* * *

The One Missing from the Room

There’s no business architect in this room.

A business architect isn’t the one who knows how agents work. They’re the one who knows why the company exists - and whether the orchestration supports that why, or quietly erodes it.

They’re not in the room because nobody invited them.

* * *

Agents orchestrate processes.

Processes are orchestrated by rules.

Rules are orchestrated by values.

Values are orchestrated by purpose.

* * *

And that - more than any hallucination, more than any misaligned prompt - is the real risk of agentic AI.

Not that the orchestra plays wrong notes.

That the orchestra plays the wrong piece.

* * *

Purpose cannot be orchestrated by anyone but you.

* * *

Author’s Note

The characters in these stories are real thinkers whose published work inspired these dialogues. The conversations are fiction. The ideas build on their foundations — the synthesis is mine.

Harrison Chase, Andrew Ng, Subbarao Kambhampati, and Karim Lakhani are publicly accessible through essays, lectures, and books. Ana Kovac is a fictional customer — and every customer your system knows only through data.

Everyone’s chasing AI value.

Few know what it actually takes.

* * *

In Kranj, a city that isn’t Ljubljana, but close enough to know all the trends, and far enough to watch them with a critical eye - there was a company.

An ordinary company.

With a CRM that tracked everything. A dashboard showing conversions, bounce rate, and customer lifetime value in real time. An NPS survey completed each month by twelve percent of customers. Annual reports showing growth.

All the data.

Not a single customer anyone knew by name.

* * *

And then there was a meeting.

An ordinary meeting. PowerPoint. Four agenda items. Coffee going cold.

Someone was presenting a new AI initiative. Process automation. Cost optimization. Productivity up thirty percent. The slides were polished. The numbers were convincing.

And then, nobody remembers who, someone asked:

“But what does Ana think about this?”

A brief silence.

Everyone looked at the dashboard.

The dashboard had no answer.

Not because Ana wasn’t in the system. She was. Loyalty score 94. Eight years. Every month. A promoter.

But nobody in the room knew who Ana was.

Nobody knew where she lived. What worried her. What made her happy. Why she stayed. What would make her leave.

They had all her data.

They didn’t know her.

* * *

Aleš was sitting in the corner.

He was quiet that day — which is unusual for him, since he normally has an opinion about everything. But that question — “What does Ana think?” — landed differently.

Not as a strategic problem.

As a personal memory.

Because once - before he became a strategist, before he became a business architect, when he was still a business analyst with a notebook and too many questions - he sat across from real customers. Listened. Took notes. Understood that value doesn’t come from the system. It comes from the person who uses it.

And that Ana, every Ana, knows something no dashboard will ever show.

That morning in Kranj, Aleš thought:

“What if we built a room where Ana isn’t a data point, but a voice?”

* * *

The AI Value Boardroom was born from that question.

Not as an answer. As a space where the question is allowed to be asked.

Every topic. Multiple boardroom perspectives. And Ana, always at the top.

Not as a goal at the end of the value ladder.

As a compass that tells us whether we’re heading in the right direction at all.

* * *

The orange dot on the AI value scale is not there by accident.

It is the voice of the customer.

It is Ana.

It is the reason we’re building any of this.

* * *

The first story awaits: “Orchestra Without a Conductor”

Five stories first. Then frameworks, methodology, and assessments. The stories are the foundation. Both matter.

AI adoption is increasingly becoming AI transformation, and that puts the accountability squarely in the boardroom.

The question in that room has shifted. It is no longer whether a company uses AI, but whether it knows how to turn it into value and competitive advantage. Most companies do not yet know how.

And AI value never lives in one place. It runs through strategy, through people, through the customer, through AI itself, and through what the whole economy is becoming. Most publications follow one of these threads. This one is built where they converge.

The AI Value Boardroom is a publication about the value AI creates, or fails to create, when the people who should be shaping it are looking through only one lens. Five dimensions. At least six roles. Built on AIVaaS™, a methodology that places AI ambition at the centre of business strategy, not alongside it.

Every piece in this publication sits somewhere in that intersection. Not every piece covers all five dimensions. But across the whole, you will find the full picture that no single view can provide.

You will also meet ‘Ana’. 🟠 The orange dot in the logo. The customer who is rarely in the room where AI decisions are made. She is the reason any of this matters.

The room is open. Every seat has a perspective worth bringing.

→ Understand the room: About

→ Find your seat: For Your Role

→ Follow the logic: The Spiral

The first stone has been thrown. The first circle is yours

.