The Financial Conduct Authority (FCA) has taken a significant step toward shaping the future of artificial intelligence (AI) in retail financial services with the publication of the Mills Review on 6 July 2026. This comprehensive report outlines how AI is set to transform the sector and sets out a roadmap for regulators, industry players, and policymakers to prepare for this shift. The review highlights both opportunities and risks associated with AI adoption, emphasizing the need for proactive governance, consumer protection, and regulatory clarity.
The Mills Review was commissioned in January 2026 by the FCA and led by Sheldon Mills, an executive director at the authority. Its primary goal was to assess the implications of advanced AI on consumers, retail financial markets, and regulators. To gather insights, the review included a survey of over 5,000 UK retail financial services consumers, offering a unique perspective on how individuals perceive and interact with AI-driven tools.
The report is nearly 150 pages long and contains extensive findings across multiple sectors, including retail banking, savings, investments, consumer finance, insurance, pensions, and payments. These insights are crucial for firms operating in these areas, as they provide a clear indication of where AI is likely to have the most impact over the next decade.
The AI Autonomy Spectrum: A Framework for Understanding AI Evolution
One of the key contributions of the Mills Review is its introduction of the AI autonomy spectrum, a classification system that helps explain how AI will evolve from an assistive tool to an autonomous agent. This framework outlines five levels of agent autonomy, which the review expects firms to move through over time:
- Level 1: Human as operator – AI supports on demand (e.g., summarizing product terms).
- Level 2: Human as collaborator – human and AI plan and act together (e.g., AI compares savings products with user input and refinement).
- Level 3: Human as consultant – AI leads, human gives guidance (e.g., AI builds a switching plan and asks the users for preferences).
- Level 4: Human as approver – AI acts, human signs off key steps (e.g., AI prepares a product application for approval by the user).
- Level 5: Human as observer – AI executes, human monitors (e.g., AI executes approved actions, the user reviews activity on a log).
This spectrum is not just theoretical—it reflects real-world adoption patterns and provides a useful tool for firms to understand how their use of AI might evolve. As firms move up this scale, they will face increasing challenges in terms of accountability, transparency, and regulatory oversight.
The Path Forward: Seven Key Recommendations
The Mills Review outlines seven key recommendations for the FCA Board and Executive, each aimed at addressing specific risks and opportunities associated with AI adoption:
1. Secure and Adapt the Regulatory Perimeter
The review calls for a review of how general-purpose Large Language Models (LLMs) operate outside the current regulatory perimeter. This includes examining advice boundaries, business tests, and financial promotions rules to ensure that LLMs are not circumventing existing regulations.
2. Strengthen System-Wide Coordination and Oversight
To address cross-border risks and opportunities, the FCA should enhance coordination with domestic authorities and international partners. This is especially important in the absence of a dedicated AI regulator or legislation.
3. Monitor the Transition to Autonomous Models and Adapt Regulatory Frameworks
As AI becomes more autonomous, regulators must adapt their frameworks accordingly. The FCA, alongside the Prudential Regulation Authority (PRA), should support better approaches to AI model risk management and work with firms, developers, and researchers to improve explainability, assurance, and governance.
4. Scale Up the FCA’s AI Lab
The review recommends expanding the FCA’s AI Lab to assess AI models used in financial services. This would help regulators anticipate how new technologies could shape the sector before they are widely adopted.
5. Enable the Foundations for Agentic Finance
To support safe adoption of autonomous AI, the FCA should lead the development of a trusted framework that clarifies how agents can be authorized, identified, and held accountable. This includes setting clear expectations around consent, identity, control, and liability.
6. Build and Adopt an AI-Enabled Agentic Supervisory Model
The review proposes the creation of an AI-enabled agentic supervisory model, which would allow regulators to detect cross-firm patterns, emerging harms, and system-wide risks more effectively. This could fundamentally change how compliance, supervision, and enforcement are conducted.
7. Develop a Trusted Public-Interest AI-Enabled Financial Capability Service
The FCA should take a proactive role in developing a free, inclusive, and trusted AI-enabled financial capability service. This would provide consumers with access to reliable information, guidance, and support from trusted sources.
Quantum Computing: A Related Strategic Uncertainty
While not the focus of the Mills Review, quantum computing is identified as a separate but related strategic uncertainty. The review warns of the “harvest now, decrypt later” risk—where data collected today could be decrypted once a sufficiently powerful quantum computer becomes available. Firms are urged to identify vulnerable cryptography, build in crypto-agility (the ability to switch cryptographic methods), and plan migration to post-quantum standards.
Interesting Reads
The Road Ahead: A Call for Proactive Governance
The Mills Review is a comprehensive attempt to map the future of AI in retail financial services. It highlights both the transformative potential of AI and the risks it poses if not properly managed. As the FCA implements its recommendations, we can expect significant regulatory changes over the coming years.
Given the rapid pace of AI adoption and the increasing level of autonomy in the sector, now is an ideal time for firms to undertake a board effectiveness review. Leadership teams must stay informed about AI developments and ensure they have the knowledge and tools to manage both opportunities and risks effectively.
In summary, the Mills Review serves as a crucial guide for navigating the complex landscape of AI in financial services. It underscores the importance of proactive governance, consumer protection, and regulatory clarity as the sector moves toward an increasingly autonomous future.
Original Source
This article is based on publicly available reporting. For the complete original story, visit the publisher’s article.
Leave a Reply