Full Breakdown
UK Financial Regulator Calls for AI Regulation After Landmark Mills Review
7/6/2026, 8:06:49 PM
Mills Review Findings on AI in Financial Services
The Financial Conduct Authority (FCA) published the “Mills Review” on 6 July 2026, a study commissioned by the FCA Board and led by executive director Sheldon Mills. The review examines large-language models such as OpenAI’s ChatGPT, Anthropic’s Claude and Google’s Gemini, and recommends that the FCA decide within the next three to six months whether to “secure and adapt” the regulatory perimeter to cover these general-purpose AI tools that currently sit outside existing financial-services rules.
Context: AI Adoption and Consumer Trust
A global survey cited by the review shows 81 % of financial firms are using AI, with 40 % at advanced stages of scaling. In the UK, firms are moving AI from back-office automation into customer-facing roles such as complaints handling and investment guidance. The FCA’s own research indicates 11 million adults (about one-fifth of the population) would consider using AI that can act autonomously within preset goals, while more than a quarter of UK consumers already trust ChatGPT, Claude or Gemini for financial advice despite limited awareness that regulated protections do not apply to those services.
Key Risks Identified
The review flags several systemic hazards:
- Concentration risk from reliance on a small set of technology providers, creating correlated behaviour, herding and common points of failure across the financial system.
- Potential blurring of regulated advice when chatbots deliver personalised recommendations.
- Amplification of fraud, cyber-security threats, consumer harm and market concentration as AI becomes more pervasive.
Official Statements from Regulators
FCA Chair Ashley Alder emphasized the need to keep pace with a rapidly evolving environment and to apply the FCA’s principles-based, outcomes-focused approach to AI. Sheldon Mills highlighted AI’s capacity to transform financial services by 2030, noting both significant opportunities and emerging risks. The Bank of England’s deputy governor Sarah Breeden warned that existing supervisory frameworks were not designed for autonomous agents and that relying solely on a human-in-the-loop is unrealistic. The FCA also noted it is the first regulator worldwide to conduct a dedicated AI impact study, though it is not legally bound to implement the review’s recommendations.
Industry and Legal Perspectives
Jonathan Herbst, global head of financial services at Norton Rose Fulbright, described the review as a “big question for policymakers” rather than an immediate crackdown, stressing that the issue will become more pressing as AI adoption accelerates. The review recommends expanding the FCA’s authority over “critical third parties” such as AI firms and cloud providers, and adopting an internal AI-enabled supervisory model. Representatives from OpenAI, Anthropic and Google declined immediate comment.
Verbatim Quotes
- “We need to keep pace with a rapidly changing environment and the principles-based, outcomes focussed approach we’ve taken on AI” — Ashley Alder, Chair, FCA
- “That's a big question for policymakers and one that will only become more pressing as AI adoption accelerates,” — Jonathan Herbst, Global Head of Financial Services, Norton Rose Fulbright
- “Our frameworks werenot built to contemplate autonomous agents, and relying on a human in the loop for all agent actions is unlikely to be realistic,” — Sarah Breeden, Deputy Governor, Bank of England
- “It is an arms race,” — Sheldon Mills, Executive Director, FCA (in remarks to the Financial Times)
- “While AI has the potential to improve access, personalisation and efficiency, it could also amplify risks associated with fraud, cyber security, consumer harm and market concentration.” — Sheldon Mills, FCA review
Conflicting Consumer Trust Figures
The review cites that more than 25 % of UK consumers trust LLMs for financial advice, whereas a separate FCA-commissioned survey reports that 20 % (? 11 million adults) are “open to using AI” for personal finance decisions. The two metrics reflect different question wording and thus produce a modest discrepancy in reported consumer confidence.
Next Steps and Timeline
The FCA will deliberate on the review’s seven recommendations, including expanding its AI Lab, creating an AI-enabled supervisory model, and strengthening oversight of critical third-party providers. A follow-up review on consumer-harm risks is slated for within six months, and the regulator aims to decide on any perimeter adjustments within the next three to six months.
