Is it time to mandate an AI lead in every government organisation?

Since 2024, every US federal agency has had to name a Chief AI Officer. Australia followed suit this year, requiring every agency and department to appoint one by July 2026, and 104 of 106 met the deadline.

The UK has no such requirement. With a new Minister for AI, a Director General for AI in the Cabinet Office and a Government Chief AI Officer in the Government Digital Service, it’s a fair moment to ask whether every Whitehall department and public body should have one too.

Same ambition, different method

The US mandate is explicit about its purpose: to advance America’s ‘global AI dominance’ and to ‘buy American’. Chief AI Officers are part of the demand side of an industrial strategy. That makes sense for the country that is home to most of the big AI companies and wants to stay that way.

Australia’s emphasis is different. It first required every agency to name an official accountable for using AI responsibly, then added Chief AI Officers to ‘lead the transformation’, so that the public service gets real value from AI and keeps public trust.

The UK wants all three. It’s AI Opportunities Action Plan aims to make the UK ‘an AI maker, not an AI taker’, with government buying ‘as both its largest customer and as a market shaper’, and for AI to transform public services as much as Australia does. And as the AI Playbook for the UK Government makes clear – it wants that done safely.

What’s different is the method. The US has form here: Congress legislated for agency Chief Information Officers in 1996 and Chief Data Officers in 2019. Chief AI Officers came by a binding White House directive. Australia used a mandatory whole-of-government policy with a fixed deadline. The UK has led from the centre and set expectations but no specifics on leadership requirements.

Does that leave the UK behind? Not necessarily. In both US and Australia many appointees simply added the title to an existing job. Australia got almost every agency to comply within eight months, yet its Finance Minister still told parliament AI leadership across agencies is ‘patchy‘. And while the US reports on compliance with the directive, there is little evidence yet on whether appointing Chief AI Officers in itself has accelerated the pace or scale of responsible AI adoption.

The UK picture

So what does the UK look like without a mandate? Across 86 government departments and executive agencies, I found 11 senior posts clearly responsible for leading the organisation’s own use of AI, judged by their job titles. One is vacant. The count relies on what organisations publish, so it is an indication, not a census.¹ But if there were many more, I suspect they would be easier to find.

Where they sit reveals something about how AI is being treated too: mostly as an extension of the data job. Eight of the eleven put AI and data in the same title, such as Chief Data and AI Officer or Head of Data and AI. Grade is also revealing. Most are at deputy director or grade 6 level. And none is on an Executive Team. There are three notable exceptions where departments have dedicated senior resource. HMRC, MHCLG and MoJ each have a director leading on just AI – not part of a wider portfolio. Where they report into is interesting too. HMRC’s Chief AI Officer reports into the CDIO, MHCLG Director of AI reports into the COO and the Director led Justice AI Unit sits inside MoJ’s service transformation group. In my article on whether CDO reporting lines make a difference I argued there is no ‘right’ place for a Chief Data Office – for now I’d argue that’s not the case for AI, being so new the closer they are to those with the power to lead over transformation the more effective they will be.

A name on the door

So does having a named AI lead matter? In my experience, it can certainly help. Organisations make faster progress when one person has authority to set strategy, join the dots across teams, be held accountable for results.

But that can still be true without an individual having “AI” in their job title. I know of at least one department where AI adoption is being driven by the Director of Transformation as part of a wider agenda. Arguably this is a more responsible and sustainable approach.

The real question is not what someone is called. It is whether they have the authority, relationships and organisational backing to influence the decisions that determine whether AI succeeds: investment, governance, security, workforce change and business transformation. Those levers rarely sit with a single individual.

That is the limitation of mandating a title. It can create visibility and accountability, but it cannot on its own create authority. You can put a name on the door, but the ability to drive change still depends on access to the levers of power around the board table.

Building on what we started

So where does that leave us? The US and Australian experience suggest there may be benefits in making AI leadership at least visible. Naming a senior individual creates a focal point for accountability, signals organisational intent and makes it easier for staff, suppliers and stakeholders to know who is responsible.

The UK should not ignore their example. If AI is becoming a core part of how public services operate, every department and public body should be clear about who is leading its adoption. Responsibility should not be hidden in organisation charts or spread so widely that nobody can identify who ultimately owns the agenda.

But that is not the same thing as mandating a Chief AI Officer.

As we have seen, many organisations make progress without a dedicated AI title. What matters most is not what someone is called, but whether they have the authority, relationships and organisational backing to influence the decisions that determine success: investment, governance, security, workforce change and service transformation.

The UK has already recognised this principle. The Roadmap for modern digital government says every central and local government organisation should have a digital leader on its executive committee and a digital non-executive director on its board by December 2026. That ambition matters because it places responsibility close to the decisions that shape organisational priorities and resources.

The next step should be to apply the same logic to AI. Not by prescribing a single role or reporting line, but by setting a clear expectation that every organisation identifies a visible senior lead for responsible, impactful AI adoption, is transparent about who that person is.

The real shift then comes when that person is given them the authority and levers to ensure AI adopted safely, at pace, and used to change how public services deliver outcomes, not just generate more output.

¹ The analysis covered 86 ministerial departments, non-ministerial departments and executive agencies listed on GOV.UK in September 2026. Organisations were counted where a named individual’s job title explicitly identified responsibility for leading AI across the organisation. Evidence was drawn from published senior staff data, GOV.UK, job advertisements and the trade press. As the method captures only formally designated roles, the figure should be treated as a minimum estimate.

Katharine Purser is the founder of Kompass Consulting. She is currently working with MHCLG on AI Strategy.