What happened
The FCA published the Mills Review on 6 July 2026 — described as 'the first work of its kind initiated by a regulator globally' — setting out how AI could reshape UK retail financial services by 2030 and beyond. Led by FCA executive director Sheldon Mills, drawing on 140 written submissions and a 5,026-person Yonder consumer survey, the review introduces a novel 'AI autonomy spectrum' (operator, collaborator, consultant, approver, observer) to map how regulatory risk shifts as firms delegate more to AI agents. Its headline proposal is an 'Agentic Supervisory Model' — AI-enabled tools letting the FCA monitor outcomes across firms in near real time — alongside six other priority recommendations, including scaling the FCA's AI Lab and securing the regulatory perimeter around Big Tech/hyperscaler dependency. The review concludes existing frameworks (Consumer Duty, Senior Managers Regime) remain fit for purpose and recommends no new AI-specific rules, but finds a fifth of UK adults (11 million) are already likely to use autonomous AI for financial decisions.
Why it matters
This is the first comprehensive regulator-commissioned assessment of agentic AI's systemic impact on a major financial sector globally, and its recommendations — especially the Agentic Supervisory Model and perimeter extensions to AI/cloud providers — will shape how financial firms design AI governance, model risk management, and accountability structures ahead of 2030.
Action needed
Brief the board and risk committee on the AI autonomy spectrum and assess where the firm's current AI deployments sit on it; prepare for potential FCA perimeter extension to AI/cloud dependencies and monitor Board/Executive decisions on the seven priority recommendations.