Regulator wants ChatGPT to flag any time it gives money advice to Brits
Key Points
- FCA calls on AI providers to add clear financial advice disclaimers
- Four in five young, less-experienced investors have used AI to invest
- Only 40% know there is no recourse if AI advice goes wrong
- Quilter welcomes the move but warns against overly prescriptive rules
- FCA flags firms excluding vulnerable customers from AI journeys
The Financial Conduct Authority (FCA) says it will require popular AI models like Claude, ChatGPT, and Geminin to issue a warning to users when providing financial advice. This comes as a growing number of Brits use the tools to ask about investment and money decisions.
The FCA’s data shows that 80% of ‘less-experienced investors’ aged 18-40 have used AI to help with investing, but only 40% of consumers correctly recognise that there is no recourse if something goes wrong after acting on advice from a general-purpose AI tool.
In response the regulator said that it’s answer is not simply to expand its remit, but to make the existing boundary clearer.
Jessica Rusu, Chief Data, Information and Intelligence Officer at the FCA, set out the call in a speech on Monday (19 October). “Through InvestSmart and the FCA Firm Checker, we already help consumers understand the protections they have. But warnings may be more effective at the point where AI-generated financial information is delivered.
“So we are calling on AI providers to help consumers understand the limits of financial information and advice generated by their models through clear, proportionate and consistent disclaimers.”
The FCA also said it would not build a new public-interest financial capability tool, which the Mills Review had recommended in July.
It will instead build on existing work, including the first TechSprint from IOSCO, the international body for securities regulators, which runs on the FCA’s AI Lab.
Wealth manager Quilter has welcomed the move. “General-purpose AI is increasingly influencing financial decisions before someone ever reaches a regulated firm, so consumers need clarity about what these tools are providing and what protections they have,” said Sam Christopher, Proposition Director at Quilter.
The firm said its own research found that 67% of non-advised consumers are comfortable using AI for at least one financial task.
It added that 40% would use AI to understand tax rules, and almost a quarter would use it to recommend investments.
Christopher warned against making the response too prescriptive while the technology is still developing.
“Regulation needs to be able to evolve alongside the technology rather than trying to anticipate every possible use case today,” said Christopher.
New guidance for banks and lenders
The FCA also published examples of good and poor practice for regulated firms that use AI with their customers.
It found that some firms excluded customers with signs of vulnerability from AI journeys by default, without offering them another route to support.
Other firms overstated what their AI systems could do, or launched automated customer journeys without enough testing.
One firm built a live list of vulnerability signals that its chatbot uses to flag customers who need extra help to human advisers.
The FCA said firms should let customers choose between AI and human support, and switch channels without repeating their information.
Rusu said firms must also report significant AI failures and near-misses under the FCA’s existing incident reporting rules.
The FCA plans to use AI tools to oversee the 60,000 anti-money laundering firms joining its remit in 2028.
A joint survey from the FCA and the Bank of England, due next week, shows that 86% of financial firms now use AI
