The CAIO is here, but the industry has not settled what AI ownership actually means. Here is the role that settles it.

The last article ended on a deliberate tension.

It argued that the discipline that matters most in the Human Intent era is the ability to define intent clearly enough that humans and systems can act on it reliably. Three capabilities. Define, translate, judge. Across every function, your value moves upstream, from doing the work to defining what the work should achieve.

And then, in the closing section, I put a limit on the argument.

Practitioner-level capability does not flourish in a vacuum. You can shape clay carefully, but if no one is tending the kiln, your work will not survive the firing. If intent is genuinely the unit around which Human Intent organisations are built, then someone at the top has to own it. Not as one priority among many. As their primary responsibility.

That role, I said, does not yet exist by name.

This piece is about why it needs to. What it looks like. Why the existing C-suite, even after several years of restructuring around AI, still cannot quite reach it. And what changes when intent moves into the boardroom.

A Familiar Pattern

Two articles ago, in The Cloud Said It First, I argued that AI is repeating cloud computing’s organisational arc on a compressed timeline. Sceptical reception, native-first adopters pulling ahead, platform layer emerging, governance gaps opening up faster than the disciplines to close them. The lesson from cloud was uncomfortable: technology arrives faster than the structures to govern it, organisations get hurt before disciplines emerge, and the ones who build governance in early pull ahead in ways that compound.

That argument is harder to make now than it was even six months ago. Not because it is wrong, but because the evidence has accumulated to the point where it barely needs making. Forrester’s Jeff Pollard, in a recent TechTarget piece on agentic AI governance, described the shift in unusually direct terms: agentic systems are “the first time we have widely deployed software in our environment that has an intent, a goal, and has the ability to go do something without us explicitly telling it what to do.” Deloitte’s 2026 enterprise survey put the gap in numbers: 74% of companies plan to deploy agentic AI moderately or extensively within two years, but only 21% report a mature agentic AI governance model. Independent analyses, including Camunda’s 2026 State of Agentic Orchestration & Automation report, put the share of agentic AI pilots reaching production at around 11%. The rest stalled, primarily through governance breakdowns rather than technical failure.

These are individual data points from individual sources, and I want to be careful not to overstate what they collectively prove. They do not yet amount to industry consensus. What they do amount to is convergence, a growing number of credible voices arriving at a similar diagnosis from different angles. Governance is lagging. The lag is structural, not incidental. And the structural answer has not yet been named.

That convergence is the soil in which this piece is planted.

What the Industry Has Tried

The most prominent response so far is the Chief AI Officer.

The role has scaled rapidly. IBM’s 2026 CEO study, surveying 2,000 CEOs across 33 countries, found that 76% of organisations now have a Chief AI Officer, up from 26% just a year earlier. The US federal government, through OMB Memorandum M-24-10, now requires every federal agency to designate a CAIO. The EU AI Act has created de facto demand for a named executive accountable for AI compliance. By any measure of corporate adoption, the CAIO has arrived.

It has also, almost immediately, struggled to settle.

The literature on the role is unusually candid about this. IBM’s own write-up acknowledges that “given its newness, its responsibilities are not well-defined.” Multiple analyst pieces describe the relationship between the CAIO, the CIO, the Chief Data Officer, and the CISO as a “complex power struggle over who truly owns AI.” In a thoughtful February 2026 piece, CIO.com observed that the role may eventually dissolve back into existing leadership: “paradoxically, the long-term success of this role may reduce the need for its existence.”

I do not think the CAIO role is wrong. It has done useful work. It has given organisations a named owner for AI strategy at a moment when one was urgently needed. It has translated regulatory demands into executive accountability. It has accelerated experimentation and helped companies move from isolated pilots to portfolio thinking.

But it does not, and structurally cannot, fill the gap I am describing.

The CAIO is defined around AI. Strategy for AI. Governance of AI. Value creation through AI. Risk management for AI. The role’s centre of gravity is the technology and its responsible deployment. Read the role descriptions across IBM, PwC, Gartner, and the executive search firms recruiting into it, and you will find variations on the same theme: the CAIO owns the AI agenda.

That is a real and important accountability. It is also not the same as owning intent.

The Distinction That Matters

Owning AI means being accountable for the technology, its selection, its safe deployment, its alignment with business goals, its ethical operation, its compliance with evolving law.

Owning intent means being accountable for the meaning that the technology is being asked to act on. The clarity of the purpose. The fidelity with which strategic ambition translates into specifications that humans and systems can execute against. The integrity of intent as it flows from boardroom to domain to project to keystroke. The honesty of the feedback that travels back the other way.

These are different jobs.

A Chief AI Officer can be excellent at their work and the organisation can still ship the wrong thing, beautifully. The systems behave responsibly. The governance frameworks are intact. The compliance documentation is impeccable. And the AI-accelerated execution delivers exactly what was specified, while the specifications quietly fail to reflect what the organisation actually meant.

The CAIO is not structurally positioned to catch this. Their accountability stops at the AI layer. The fidelity of intent above that layer, whether the strategic purpose was articulated clearly enough to be acted on faithfully, sits outside their remit by design.

The same is true, in different ways, for the other seats around the table.

The CIO, in its traditional sense, owns information systems and IT operations. The role is broadening, helpfully. Recent forecasts have CIOs evolving into more strategic, business-facing leaders. But the centre of gravity remains technology infrastructure and its operation.

The Chief Data Officer owns data quality, lineage, and access. Essential foundational work. Not intent.

The CISO owns security and threat modelling. Increasingly, that includes agent identity and runtime monitoring. Vendors like Zenity have built whole platforms around what they call “intent-aware detection” at the agent layer. Important work. Different layer.

The Chief Transformation Officer is typically time-boxed and programme-shaped. The Chief Innovation Officer spots opportunity. The COO runs operations. The CFO owns the numbers. The General Counsel interprets the rules.

Each of these roles intersects with intent. None of them owns it.

The result is that intent, the artefact that everything else is now organised around, has no executive owner. It is everyone’s concern and nobody’s accountability. And as the cloud article argued, accountability that defaults to “everyone” is the same as accountability that defaults to nobody, which is the same as accountability that defaults to whoever happens to be in the room when something goes wrong.

That is not a stable arrangement.

The Title

I want to give the role a name, knowing the name will be the most contested thing about it.

Chief Intent Officer.

The same three letters as the historical CIO. A different word in the middle. A succession rather than a replacement, in the same way the 1980s Chief Information Officer was a succession to the Data Processing Manager, not an erasure of what came before, but a recognition that the strategic centre of gravity had moved.

The phrase is not a corporate invention. It draws from a much older lineage. In military doctrine, commander’s intent is the formal name for the discipline of expressing the purpose of an operation clearly enough that subordinates closer to the action can exercise initiative when conditions change faster than orders can be issued. The FBI’s Law Enforcement Bulletin describes it as “a clear and concise expression of the operation’s purpose and the desired end state.” The US Coast Guard, the British Army, and emergency-services manuals across multiple jurisdictions use the same construct, refined across decades of high-stakes operational reality.

The doctrine exists because in fast-moving situations, top-level decision-makers cannot incorporate new information into a formal planning process and redirect people to action within a reasonable timeframe. Leaders provide intent so people closest to the scene can adapt plans and exercise initiative when the original plan no longer fits.

That is, almost word for word, the problem agentic AI has just deposited on the boardroom table.

The Chief Intent Officer is not borrowing a metaphor. They are inheriting a discipline that has been refined for decades in environments where getting intent wrong gets people killed. That is heavier provenance than most C-suite titles can claim, and it is sitting unused.

What the Role Actually Does

Three responsibilities, which mirror at the organisational level the three practitioner-level capabilities I described in the last article: define, translate, judge.

Define. The Chief Intent Officer is accountable for the clarity of organisational intent. Not for setting strategy, that remains the CEO’s job, but for the discipline with which strategic ambition is articulated as intent that can survive the journey to execution. Where executive teams describe activities, the CIO presses for outcomes. Where they describe assumed solutions, the CIO presses for the underlying purpose. Where intent is fuzzy, constraints unstated, success criteria aspirational, the CIO is the named owner of fixing it. This is closer to executive coaching practice than to traditional management oversight, and the role demands the skills to match.

Translate. The Chief Intent Officer owns the architecture by which intent flows through the organisation. From organisational intent, to domain intent, to project intent, to the specifications that humans and agents act on. Each handoff is a place where intent leaks. The CIO is responsible for the integrity of that flow, the structural mechanisms that prevent boardroom purpose from arriving at the keystroke as something the boardroom would not recognise. This is the executive analogue of the practitioner translation skill, scaled to the architecture of the enterprise.

Judge. The Chief Intent Officer is accountable for the bidirectional flow of intent and reality. Strategic intent travels down. Execution reality travels back up. Both must remain honest. The CIO holds the human gates that decide which decisions delegate to agents and which stay with humans, where the autonomy boundaries sit, when intent itself needs to be revised in light of what execution has revealed. This is closer to the senior judgement of a board director than to traditional management. It demands intellectual honesty about whether the organisation is achieving what it set out to achieve, or just achieving something it once asked for.

These three responsibilities do not fit neatly onto any existing seat. They are the seat.

What Changes When Intent Moves Into the Boardroom

The most interesting effect, in my experience, is what the role dissolves.

The current AI-era turf war between the CAIO, the CIO, the CDO, and the CISO exists because all of them are trying to own AI. The literature is saturated with “how do these roles avoid conflict” pieces precisely because the question is contested by design. Each role has a legitimate claim. None of the claims aggregate into clear accountability.

Once intent is the centre of gravity, the turf war loses its energy. The CAIO owns AI strategy and value creation, contributing to intent fidelity at the AI layer. The CDO owns data quality and lineage, contributing at the data layer. The CIO-traditional owns information systems and operations, contributing at the platform layer. The CISO owns security, contributing at the runtime layer. Each contributes upward to the integrity of intent. None of them is asked to own something they were never structurally positioned to own. The Chief Intent Officer governs the artefact that all of them serve.

The boardroom conversation also shifts. Different questions get asked when an Intent role is in the room. Not “is the AI compliant,” that question is still asked, but elsewhere. The questions that move to the centre are the ones the convergence I described earlier has been struggling to ask cleanly:

• Have we articulated what we actually want with enough precision that the systems we are deploying can pursue it faithfully?
• Where is intent leaking between strategy and execution?
• Are we measuring what was achieved against what was meant, or only against what was specified?
• When execution reveals something we did not anticipate, who has the authority and the discipline to revise our intent rather than just our delivery plan?

These are not new questions. They have been asked, in different vocabularies, by good leaders for decades. What is new is that the apparatus surrounding them, the methodologies, the structures, the executive accountabilities, was built for a world where execution was the constraint and intent could remain implicit. That world is now ending. The questions have to come into the open. Someone has to be named accountable for asking them, hearing the honest answer, and acting on it.

A Note on Where the Role Sits

I am deliberately not prescribing whether the Chief Intent Officer is a new seat or an evolution of an existing one. The right answer will vary.

In some organisations, the existing CIO will recognise this is where the role has been heading and pull the accountability into themselves, perhaps with a new title or perhaps without one. In others, the CAIO will mature beyond AI strategy into a broader intent governance remit, taking on the new word as the technology becomes infrastructure. In some, the role will be created fresh, alongside the existing seats, with explicit accountability for the gap none of them currently fills. In smaller organisations, the CEO will keep it close.

What matters is not the org chart. It is whether the question I asked at the top of this piece has a single clear answer.

Who, by name, is accountable for the fidelity of intent from strategic purpose to executed action?

If the answer is “everyone” or “nobody” or “well, sort of the CEO and sort of the CTO and sort of the CAIO,” the gap is real. The role has a name now, whatever you choose to call it locally. The convergence I described at the start of this piece is closing in around it. The organisations that name the accountability early will, on the cloud parallel, pull ahead in ways that compound.

The discipline begins with practitioners. It does not survive without leadership. And leadership, in the era we are now entering, looks less like managing the work and more like governing the meaning the work is meant to serve.

Govern Intent. Delegate Execution.

That is what the next CIO is for.

This is the sixth article in the Human Intent series. Earlier articles traced the shift from managing work to governing intent, the strain on existing methodologies, the end of Taylor’s century, the dissolution of work as the organising unit, and the practitioner-level capabilities that replace execution competence as the source of professional value. The cloud parallel runs alongside the series as a reminder that we have seen this pattern before, with consequences that are now well documented.

Frequently Asked Questions

Next Article ->

Human Intent – The Human Intent Stack

The Human Intent Stack: how intent moves from human purpose to machine execution, and the world-readiness test most organisations are not yet running.