TL;DR
The FCA has launched a long-term review into AI and retail financial services, examining how emerging technology could affect consumers, markets and firms through 2030 and beyond. The regulator is seeking industry input before recommendations go to the FCA Board this summer. Consultation closes 24 February 2026.
Designing for the Unknown
Sheldon Mills, leading the review, framed the challenge clearly: “The real challenge in regulation isn’t dealing with what we already understand—it’s preparing for what we don’t.” The review aims to help the FCA continue shaping AI-enabled financial services while maintaining its outcomes-based, technology-neutral approach.
The context is significant. Lloyd’s 2025 survey found one in three customers use AI weekly to manage their money. The FCA’s Supercharged Sandbox received 132 applications, selecting 23 firms at “the frontier of retail financial services.” AI in finance is no longer theoretical.
The Proxy Economy Scenario
Mills outlined a scenario where consumers increasingly use AI as an intermediary with financial firms. This progression moves from assistive AI (explaining products, comparing options) through advisory AI (nudging recommendations, suggesting switches) to autonomous AI agents acting within user-defined boundaries.
He made this concrete: “Imagine Sarah, a working parent in 2030. Her AI agent manages household finances within agreed boundaries—moving money to higher-rate savings, flagging uncompetitive renewals, even switching current accounts on her behalf.”
But agent autonomy raises difficult questions: What happens when an AI agent makes a mistake? How do consumers stay in control? What if commercial incentives quietly shape recommendations?
Regulatory Frameworks Under Strain
Current frameworks assume systems update occasionally, models behave predictably and responsibility sits clearly within firms. AI challenges all three assumptions.
“What does ‘reasonable steps’ look like when the model you rely on updates weekly, incorporates components you don’t directly control, or behaves differently as soon as new data arrives?” Mills asked. The review will examine how existing rules apply when AI systems evolve continuously and responsibility spans developers, data providers and deploying firms.
Looking Forward
The FCA wants input on opportunities and risks as AI becomes more capable, how AI could reshape competition and customer relationships, and how regulatory frameworks may need to adapt. The deadline is 24 February 2026.
Mills closed with a direct request: “Contribute. Challenge our assumptions. Tell us what we’re missing.” For UK financial services firms working with AI, this review offers genuine opportunity to shape the regulatory environment before structures become fixed.