AI in action · Regulation

The FCA Wants to 'Automate Wherever We Can', and Says a General AI Chatbot May Sit Outside Its Rules

At a House of Commons Treasury Committee hearing on October 13, 2026, the UK regulator's chief data officer said the FCA aims to automate supervision and enforcement wherever it can. Its chief executive said he is not yet convinced the regulator needs more powers, and that a chatbot giving general information is not wholly inside the regulated perimeter.

The Financial Conduct Authority supervises about 42,000 firms, and on October 13, 2026 it told MPs how it plans to do that with artificial intelligence. Jessica Rusu, the regulator's chief data, information and intelligence officer, told the Treasury Committee that the FCA's aim is "to automate wherever we can" and "to intervene at scale", moving as much supervision and enforcement work to AI as possible, according to FStech and Retail Banker International. The idea is an AI-enabled supervisory model that processes the data firms already report and flags risks earlier than human reviewers can.

On the firms it regulates, the message was the opposite of new rules. Chief executive Nikhil Rathi said he is still not convinced the FCA needs "lots more powers" until it has shown it has used its existing ones to the fullest extent. The regulator's position, as reported by dig.watch and Finance Magnates, is that existing consumer-protection and accountability rules apply to firms' use of AI, and that it has no timetable for broader automation or for AI-specific regulation.

The sharpest exchange was about chatbots. Rathi told the committee that an AI model giving general financial information, without a personal recommendation that leads to a transaction, is not wholly inside the regulated perimeter, and called where that line should sit "a very live discussion". For anyone asking a chatbot which fund to buy, that is a precise statement of what the law does and does not protect: the answer may be unregulated right up to the point where it becomes personal advice.

Rusu said the FCA would publish findings on good and poor AI practices at financial firms the following week, alongside its response to the growing use of advanced AI models. The committee itself has urged the regulator to publish its AI guidance by the end of 2026. In June, Rathi's speech "Rethinking regulation for the age of AI" had already set out the principle that the FCA would rather test tools with firms than write rules in advance.

For investors the practical point is the perimeter. A regulated adviser owes you duties that a general-purpose chatbot does not, however confident its answer. The FCA is telling you that it will police firms with AI, not that AI's answers are policed.

Sources: FStech — FCA plans to automate supervision and enforcement using AI (Oct 2026); Retail Banker International — UK's FCA seeks to automate oversight via AI; Finance Magnates (via TradingView) — FCA plans to automate supervision with AI, holds off on rules for chatbots; FCA — Rethinking regulation for the age of AI (speech, Jun 24, 2026)