What happened
At its annual public meeting (6-7 October 2026), FCA Chief Data, Information and Intelligence Officer Jessica Rusu said the FCA is taking forward Mills Review recommendations — including becoming an 'agentic regulator' (an AI-enabled supervisory model) and scaling innovation services/AI Lab — while confirming the FCA does not plan to make AI-specific rules and will instead rely on existing rules such as Consumer Duty. The FCA is publishing AI good- and bad-practice findings for firms in the coming days.
Why it matters
The UK's major financial regulator is explicitly declining AI-specific rulemaking in favour of an 'agentic' supervision posture, giving a clear compliance direction signal: firms must read AI risk through existing Consumer Duty and perimeter rules, and will face a regulator that itself uses AI agents in supervision while publishing observed good/bad deployment practices.
Action needed
UK financial firms should review the FCA's forthcoming good/bad AI practice publication, and ensure agentic-AI deployments operate under existing rules (Consumer Duty, operational resilience) rather than awaiting AI-specific rules.