CTO vs CIO vs Chief AI Officer: Who Owns AI?

Somewhere in most organisations right now, a version of the same conversation is happening. The board wants an AI strategy. The CEO wants a name against it. And three people at the executive table — the CIO, the CTO and sometimes a newly minted Chief AI Officer — are quietly wondering whose job it actually is.

The question sounds like a turf dispute. It is really a design question about decision rights. Before an organisation can answer who owns AI, it has to be honest about what each of these roles was built to own — and why AI refuses to fit neatly into any of them.

What the CTO and CIO were each built to own

The classic division of labour is straightforward. The CIO owns the systems the business runs on: ERP, finance, HR, infrastructure, security posture, the estate of vendors and licences. Success looks like reliability, cost discipline and risk control. The CTO owns the technology the business sells or differentiates with: product engineering, platform architecture, technical talent and the roadmap that turns technology into revenue.

In a product company the CTO leans commercial; in an operations-heavy enterprise the CIO often carries more weight. Both roles evolved for a world in which a given technology was either the product or the plumbing — rarely both at once.

AI breaks that assumption. A single capability — say, a model that drafts customer responses from your knowledge base — is simultaneously an internal efficiency system (CIO territory), a potential product feature (CTO territory) and a governance question (legal, risk and the board). That is why AI ownership debates feel circular: the technology genuinely spans mandates that were designed to be separate.

The case for a Chief AI Officer — and the case against

The argument for a dedicated Chief AI Officer is focus. AI adoption is not one project; it is a portfolio of use cases, a data and knowledge estate, a governance obligation and a cultural change, all moving at once. Giving that portfolio a single accountable owner prevents it from becoming everyone’s second priority.

The argument against is fragmentation. An AI officer without authority over engineering, data or budget becomes a coordinator of other people’s decisions — an evangelist with a title. Worse, a standalone AI function can become a lab that ships demos while the operating businesses carry on unchanged. I wrote about that failure mode in From Pilots to Capability: the pattern where pilots succeed and nothing changes is usually an ownership failure, not a technology one.

The deciding factor is not company size but accountability. If AI outcomes — revenue, efficiency, risk — are material enough to appear on the board agenda with numbers attached, they deserve an owner with a budget and the authority to change how work is done. If they are not yet material, a dedicated title is premature; give the mandate to the CTO or CIO explicitly instead of leaving it ambient.

Why I chose the combined route

At Webskitters we answered the question by combining rather than splitting: I moved from CTO to Chief Technology & AI Officer. The reasoning, which I set out in From CTO to CTAIO, was that separating AI from technology leadership would have created exactly the seam we were trying to avoid — one leader accountable for the systems and another for the intelligence inside them.

The combined role only works under one condition: AI accountability has to be explicit, not implied. That means named outcomes, a defined governance practice and a clear view of which workflows are being changed. A title change without that structure is rebranding.

A practical ownership test

Whatever structure you choose, run it through three questions. First, who owns the value case — can one person say which business outcomes AI is accountable for this year, and defend the numbers? Second, who owns governance — is there a named owner for model risk, data use and compliance obligations under frameworks such as the NIST AI Risk Management Framework and, where it applies, the EU AI Act? Third, who owns adoption — who is accountable when a capability works in a demo but is not used on a Tuesday afternoon?

If those three answers are three different people with no shared forum, you have found your problem, and it is not the org chart’s label. If they are one person — whether that person is called CIO, CTO or Chief AI Officer — you have an owner. The title is the least important part of the answer.

The org chart question will keep coming back as AI matures, and the honest answer may change as your organisation does. What should not change is the principle: AI is owned where the value case, the governance and the adoption meet — and it is the leadership team’s job to make sure they meet somewhere.