UK watchdog urged to review AI tools giving financial guidance

Britain’s financial regulator should examine whether chatbots and other AI systems are giving consumers financial help that falls outside existing rules, a senior official has warned. Sheldon Mills, an executive director at the Financial Conduct Authority, told the Financial Times that regulators will need stronger tools as AI spreads through savings, borrowing and other personal finance decisions.

Mills made the comments ahead of the publication of an FCA-commissioned report he has written on AI in financial services, according to the FT. He said regulators may have to use AI themselves to keep pace with how quickly the technology is changing the sector and to identify risks more effectively.

The report points to both gains and dangers from wider AI use. According to a summary seen by the FT, AI could match financial products more closely to individual needs, while also creating risks around biased outcomes, unclear pricing and tailored manipulation of consumers.

Mills’ report recommends that the FCA review the issue within three to six months. That review would look at risks from companies offering financial services beyond the regulator’s reach, as well as possible consumer harm from people using large language models such as ChatGPT, Claude and Gemini to manage money.

Research commissioned by Mills found that one-fifth of UK adults were open to letting AI models make financial decisions for them, including choices about savings or borrowing, the FT reported. Those tools are not currently covered by financial regulation in the same way as authorised advisers, and consumers would not have access to compensation if AI-generated decisions caused losses.

Mills told the FT that some firms view chatbot-based financial help as economically similar to regulated services while sitting outside the FCA’s perimeter. He raised the question of whether a conversational AI system responding to prompts is closer to financial guidance or to a recommendation, a distinction that matters because regulated firms face stricter requirements when giving advice.

The report also argues that AI could expand access to financial services now associated with wealthier customers. Mills told the FT that people earning £20,000 a year could one day receive advice comparable to services often available only to customers with £10 million in savings or assets.

One recommendation calls for the FCA to bring together public and private sector groups to build an AI-supported service offering free financial information and guidance to people in Britain. The report also says human managers should remain responsible when financial firms deploy AI agents that can carry out transactions for companies or consumers.

Fraud and cyber risks feature heavily in the report. It says AI is likely to increase threats from deepfakes, synthetic identities and personalised social engineering, and recommends using AI-based defenses to protect the financial system.

Mills also recommends expanding the FCA’s powers under the UK’s critical third parties regime, which lets regulators oversee key technology suppliers to the financial sector. The FT named Anthropic, OpenAI, Amazon, Google and Microsoft among the types of providers that could fall within that framework.

The UK government has not yet decided which major technology companies will be designated under the regime. Once designated, companies could face stronger disclosure duties, including annual self-assessments and tests of their ability to withstand severe disruption.

The report says the FCA could also seek powers under the designated activities regime, which allows regulation of specific activities without requiring the firms performing them to become fully authorised. The FCA board is expected to discuss Mills’ report before deciding how to respond.

This story draws on original reporting from Ars Technica.