UK FCA review and AI Adoption Plan target agentic finance

The FCA’s Mills Review and the UK Treasury’s AI Adoption Plan are not splashy pieces of legislation, but they quietly sketch how a mature financial system intends to live with increasingly agentic AI. The documents, as summarized by PwC, focus less on banning risky models and more on integrating AI into existing outcomes‑based regimes, with tweaks around general‑purpose tools, agent frameworks and supervisory technology. That is a very British approach: keep the rulebook, reinterpret it for a world where AI is making more of the day‑to‑day decisions.

The emphasis on “agentic finance” and “agentic payments” is notable. Regulators are starting to anticipate systems where autonomous agents move money, manage portfolios and interact with third‑party infrastructure with relatively little human in the loop. That drives recommendations for trusted frameworks around AI agents, incident reporting and third‑party assurance schemes for foundational models and cloud providers.

In the race to AGI, this matters less for speed and more for direction. If major markets like the UK can demonstrate that highly agentic financial systems can be supervised under existing principles, it will weaken arguments that AGI‑adjacent tools are fundamentally unregulatable. Conversely, if these early frameworks struggle once real failures happen, pressure will build for much stricter, possibly timeline‑slowing interventions.