Nearly a third (29%) of consumers who shopped for pensions and investments in the past 12 months used artificial intelligence, the Mills Review into the use of AI in financial services has found. It recommends adapting regulation and creating a public-interest AI-enabled financial capability service.
    

Consumers are most likely to turn to artificial intelligence for finance where decisions feel complex or high-stakes, such as pensions or debt advice, the review by Sheldon Mills for the Financial Conduct Authority has found.

A survey of over 5,000 financial services consumers conducted for the FCA by Yonder suggests that a fifth of people – around 11m UK adults – are open to using AI that can act autonomously within pre-set goals, and 13% of all consumers would be willing to give AI real-time access to banking and financial data. 

Retail finance will see major shifts from AI, Mills said, from the transformation of firm operations and evolving consumer journeys, to changes in competition and market power and to amplified fraud and cyber risks. While the risks are apparent, he argued AI can also democratise financial services. 

“Artificial intelligence will transform financial services by 2030. It creates significant opportunities for consumers, firms and the wider economy. This report sets out a roadmap for how industry regulators and government can prepare for the next phase of AI-driven change in our world-leading financial services sector,” he said. 

Mills, who stepped down from his role as executive director consumers and competition to conduct the review, recommends the FCA’s board and executive should: 

Board chair Ashley Alder thanked Mills for his work and said: “As is clear in the report, we need to keep pace with a rapidly changing environment and the principles-based, outcomes focussed approach we’ve taken on AI – relying on the Consumer Duty and Senior Managers Regime – has been critical to us doing so. The recommendations build on work the FCA has been doing – not least allowing firms to test their use of AI with us – and our own use of AI to be a smarter regulator, more efficient and effective.”

The FCA will launch an AI good and poor practice publication later this year, having engaged directly with firms. 

The Mills Review suggests that the FCA should also consider conducting a review within three to six months to examine how consumers use AI, including general purpose large language model tools, for personal financial management across savings, investments, pensions, mortgages and debt management, and the implications for competition, innovation and growth. 

This additional review should examine the risks of consumer harm, including how far its usage has or will move along the autonomy spectrum, and any impacts on market integrity and the potential for regulatory arbitrage, Mills suggested. Financial advisers had previously warned that AI tools may be providing unregulated advice with no consequences for the firms that operate them.

Consumer interactions increasingly take place through general LLMs such as ChatGPT, Gemini and Claude, the report said, “shaping how consumers interpret information, narrow options and act”, with many either unaware there is no formal redress route or falsely assuming that protections exist. 

However, it also found that consumers do not, for the most part, use AI to delegate financial decisions and that trust in AI is selective, declining for more autonomous systems. 

The report raises concerns that the hoped-for democratisation of financial support could encounter barriers if access to the most capable AI systems depends on cost, digital capability, commercial arrangements or platform choice. 

“Should frontier AI capabilities become concentrated within paid services or services with embedded commercial interests, unequal access and conflicts of interest may emerge,” the report notes. 

Mills said the FCA should therefore work with industry, government and other regulators to build a “trusted AI-enabled financial capability service that provides consumers with free access to reliable financial information, guidance and support”, saying this should include “consideration of a sovereign-style or public-interest model built on trusted financial information sources”. 

As AI becomes more personalised and able to act as agents, the financial services industry must ensure protections, accountability and redress keep pace with innovation, said Sophie Legrand-Green, head of policy: consumer protection and access at the Investment and Saving Alliance.

“Responsible AI adoption means clear rules, robust testing and strong governance. These tools must be tested against real consumer risks, including vulnerability, comprehension, bias and access to redress, before they become embedded in everyday financial decision-making,” she said.

Consultancy Aon recently placed AI output on retirement planning and human advice side by side to compare them and found that the same information fed to the same AI tool produced different recommendations and reasoning on different occasions, LLMs could easily be steered to support a preferred recommendation the custormer is set on, and calculations could be oversimplified or even inaccurate, with factually incorrect claims presented confidently.