Regulation
HM Treasury's financial services AI adoption plan puts the FCA on the hook for a review of LLM-generated financial guidance, not just agentic trading systems
The ten-point roadmap, drawn up by AI Champions from Starling Bank and Lloyds Banking Group and accepted by HM Treasury on 14 July, spans five areas from regulatory clarity to agentic payments, but its most consequential line item may be the instruction for the FCA to formally assess what happens when consumers take financial guidance from general-purpose chatbots that were never authorised to give it.
The plan did not come from the regulator itself. It was drawn up by two industry insiders, Starling Bank's Group Chief Information Officer Harriet Rees and Lloyds Banking Group's Head of AI and Advanced Analytics Dr Rohit Dhawan, and HM Treasury then accepted the recommendations largely as written. That is a different mechanism to the FCA's own recent public posture, where chief executive Nikhil Rathi has been describing a shift toward principles-based stewardship rather than new rulebooks. This plan gives the FCA specific, named deliverables with a government mandate behind them, which is closer to a homework assignment than a principle.
The sharpest edge is recommendation two: the FCA should undertake a comprehensive review of the consumer, competition and wider impacts of financial guidance-like outputs generated by general-purpose large language models. That distinction matters because the current regulatory perimeter was built around authorised advisers, not around a consumer asking a chatbot for portfolio guidance outside any advisory relationship. Asking the FCA to review that gap is an implicit admission that the existing framework was not obviously built to capture it.
The agentic payments trust framework recommendation connects to a theme this site tracks closely, agentic systems that coordinate and transact rather than just support a human's decision. Recommendation four, accelerating the Critical Third-Party regime to assess Key AI/Cloud Providers, is concentration-risk-adjacent in the same way as the index-concentration story in Alphabet's capex selloff this week and the IMF's flash-crash warning covered on this site last week: a small number of vendors now sit underneath a large share of UK financial-sector AI infrastructure, and nobody has published an official map of who they are.
Watch two things from here. First, whether the FCA's promised review of LLM financial guidance actually narrows the regulatory perimeter or leaves the gap open while merely describing it. Second, whether the Critical Third-Party assessment produces the first official named list of Key AI/Cloud Providers, since that list would be the first concentration map of exactly which vendors the UK financial system now depends on for AI infrastructure, a fact currently known only informally.
Read the original: GOV.UK / HM Treasury - HM Treasury's financial services AI adoption plan puts the FCA on the hook for a review of LLM-generated financial guidance, not just agentic trading systems. Commentary is the independent editorial view of Share Trading; the original article is credited to its publisher.