The UK Financial Conduct Authority has published the Mills Review, a report on how artificial intelligence could change retail financial services by 2030 and beyond.
The review was led by FCA Executive Director Sheldon Mills and commissioned by the FCA Board. It focuses on how AI may affect consumers, firms, markets and regulators.
The regulator said AI could change how firms operate, how consumers make financial choices, how competition works and how fraud and cyber risks develop. The FCA also said there is already public interest in using autonomous AI tools for personal finance, though many users remain concerned about control and trust.
According to JURIST, the review expects consumers and firms to hand “more financial decision-making to AI systems.”
These decisions could include agreements, transactions and other actions carried out within agreed limits. The review also said AI could help close gaps in financial advice, but it warned that the same tools could raise risks for consumers.
Consumers already trust AI for finance
The review found that one in five UK adults is already open to AI making decisions for them, mainly when choices feel complex or high stakes. It also found that about 26 percent of people trust general-purpose tools such as ChatGPT, Claude or Gemini for financial advice.
Many users may not know that these tools do not offer formal routes to recourse or the same protections as regulated financial advice.
Sheldon Mills warned that AI can raise risks linked to “bias, discrimination, exclusion, opaque decision-making, misleading or hallucinatory advice and erosion of consumer trust.”
The review said these concerns matter more when different AI models interact with each other and when users rely on them for important financial choices.
The FCA also said AI could increase risks tied to fraud, cyber security, consumer harm and market concentration. This means the regulator is watching not only advice tools, but also the systems behind payments, onboarding, compliance checks and automated customer journeys.
FCA sets seven priority recommendations
The Mills Review gave seven recommendations for the FCA Board and Executive to consider. These include securing the regulatory perimeter, improving system-wide oversight and monitoring the shift toward autonomous AI models. The review also recommended scaling up the FCA’s AI Lab, enabling agentic finance, building an AI-enabled supervisory model and creating a public-interest AI financial capability service.
The FCA said its current approach relies on existing frameworks, including the Consumer Duty and the Senior Managers Regime. FCA Chair Ashley Alder said the regulator must keep pace with a fast-changing market while allowing firms to test AI use with the regulator.
Furthermore, the review also builds on previous FCA work, including its AI Discussion Paper, AI Sprint, AI Lab, AI Live Testing service and Supercharged Sandbox supported by NVIDIA. The regulator said the review followed a survey of more than 5,000 UK retail financial services consumers.
The review adds to broader finance and AI scrutiny
The FCA’s AI review comes as regulators pay closer attention to both digital finance and automated decision-making. As reported by The Coin Headlines, the FCA recently published its final crypto regulatory framework, cutting stablecoin capital requirements from 2 percent to 1 percent of issued stablecoin value. Those crypto rules are set to take effect in October 2027.
That earlier FCA crypto framework brought exchanges, custodians and stablecoin issuers under a wider rulebook for the first time. It also requires firms to meet capital requirements, carry out annual stress tests and show how they can manage market shocks.
As we reported in May, Anthropic also launched 10 ready-to-use financial AI agent templates for banks, asset managers and insurers. Those tools were designed for research, KYC reviews, back-office finance tasks, financial modeling, valuation checks and compliance workflows.
Moreover, Anthropic stressed that users remain responsible for reviewing and approving agent work before it is sent to clients, filed or used for decisions. This point connects closely with the FCA’s concern that AI tools can support finance work but still require clear human checks and strong controls.
The FCA also plans to publish an AI “good and poor practice” document in late 2026. That update could give firms clearer examples of acceptable and risky AI use in financial services.
The Mills Review shows that the FCA is preparing for a market where AI tools may support advice, payments, fraud checks, customer support and trading-related services. The review does not ban AI use in finance.
Instead, it calls for stronger oversight, better coordination and safeguards as AI tools become more common in consumer financial decisions.
