Being named to a role and being accountable for a decision are not the same thing. The gap between them is easy to miss when the naming is public and the accountability is not, and this month's clearest illustration of that gap comes from HM Treasury itself.
In January 2026, the Treasury appointed two Financial Services AI Champions. Harriet Rees, Group Chief Information Officer at Starling Bank. Dr Rohit Dhawan, Head of AI and Advanced Analytics at Lloyds Banking Group. Both are credible, senior figures at institutions this newsletter's readers know by name. Both report directly to the Economic Secretary to the Treasury, a role currently held by Lucy Rigby KC MP. In July, they published a Financial Services AI Adoption Plan, ten recommendations spanning the regulatory framework, AI-powered financial advice, operational resilience, skills, and agentic payments.
This is genuinely useful work, produced by people with the standing to do it well. Nothing that follows is a criticism of the Champions or their plan. The point is narrower and, for a board trying to work out where its own accountability actually sits, considerably more important.
The Champions hold a voluntary advisory position. There is no Statement of Responsibilities attached to the role. No personal regulatory liability follows if a recommendation in their plan turns out to be wrong, or if a firm implements it badly. That is not a flaw in how the role was designed. It is not what the role was for. Their mandate was to accelerate safe adoption across the sector, to identify barriers and propose solutions at a national level. Accountability for individual decisions was never part of the brief, and it would be a strange thing to expect of two people who do not run the firms making those decisions.
Compare that to a fact this newsletter has covered before, because it sits on the other side of exactly this distinction. The FCA and PRA have confirmed, following consultation, that AI does not receive its own prescribed responsibility under the Senior Managers and Certification Regime. Instead, the senior manager already accountable for a business area is automatically accountable for the AI used inside it. No appointment. No press release. No plan published under their name. The accountability simply attaches to whoever already holds the role, the moment their business area starts using AI, whether or not anyone tells them so.
Set the two side by side and the shape of the problem becomes clear. One structure is visible: two named individuals, a published plan, ministerial sponsorship, media coverage. The other is invisible: no announcement, no ceremony, attaching quietly to thousands of senior managers across the regulated sector, most of whom have never been told it applies to them specifically.
A board that has watched the Adoption Plan land, with its named authors and its ten recommendations, could reasonably conclude that AI governance in UK financial services now has a visible owner, someone accountable at the national level for how this goes. That conclusion would be a mistake, and the kind of mistake that matters only when something goes wrong and the firm needs to show who was actually accountable for the specific decision in question. The Adoption Plan does not answer that question. It was never built to.
The confusion is understandable, because the two structures resemble each other from a distance. Both involve the word "AI." Both involve senior, named individuals. Both involve financial services specifically. But one is sponsorship and the other is ownership, and only one of them carries personal, evidenced consequence if a specific decision turns out to be wrong.
A senior manager's own exposure is not reduced by the existence of a well-regarded national plan, any more than a firm's compliance obligations are discharged by having read one. If a regulator asks a specific question, who reviewed this AI system's use in your business area, and when, the answer that matters is a named individual and a dated record, not a reference to the Treasury's ten recommendations. The Champions' plan can shape policy, inform sandboxes, and improve how the sector thinks about adoption at scale, all of which it may well do. It cannot stand in for the evidentiary record a specific senior manager needs to produce about a specific system, because it was never designed to hold that kind of detail, and nobody involved in producing it carries personal liability if a firm mistakes the plan for a substitute.
There is a version of good governance that treats visible, national-level activity as reassuring context, and a version that treats it as a distraction from the much narrower, much more personal question a regulator will actually ask. The second version is the one that survives scrutiny.
Three questions worth putting to your own senior manager this week, distinct from anything the Adoption Plan covers. Is there a named individual in your firm who could say, without checking, that they are personally accountable for a specific AI system, separate from general awareness that the Treasury has published a plan. Is there a dated record of that individual reviewing that specific system, rather than a general sense that the firm follows good practice nationally. And if a regulator asked for that record tomorrow, would it already exist, or would it need to be assembled once the question was asked.
The Champions were named because naming them served a purpose: visibility, sponsorship, a mandate to move the sector forward. Senior managers are accountable because the regulation says so, whether or not anyone ever names them publicly, whether or not a plan is ever published under their signature. Confusing the first kind of naming for the second is not a technical error. It is the difference between a firm that can point to a national conversation and a firm that can produce a record.
Regulatory references: HM Treasury, Financial Services AI Adoption Plan (14 July 2026); FCA/PRA joint consultation outcome on senior manager accountability for AI (April 2026); Senior Managers and Certification Regime (SYSC, FCA Handbook); Companies Act 2006, section 174.
The Roche-Review is the weekly publication of Dr Ivan Roche FRSS FRSA MInstP, Founder of Otopoetic Limited. Subscribe at roche-review.com.
