In a recent reply released this month, the Bank of England as well as the Prudential Regulation Authority (PRA) have responded to a government request from HM Treasury, the Department for Science, Innovation and Technology, and the Department for Business and Trade. The letter, signed by Deputy Governor for Financial Stability Sarah Breeden and Deputy Governor for Prudential Regulation and PRA CEO Sam Woods, addresses a 28 January appeal for a clear strategy to support safe AI innovation in financial services alongside annual updates on regulatory impact.
The authorities stress that AI, when used responsibly, can unlock meaningful innovation and economic expansion in both the financial sector and the wider economy.
Their aim is to foster an environment where responsible AI adoption boosts competition, competitiveness and growth without compromising the stability or integrity of the financial system.
They describe their stance as proactive and forward-looking, with continuous monitoring to match the technology’s rapid evolution.
The Bank and PRA have been engaged on this topic for years.
In 2022 they jointly published a discussion paper with the Financial Conduct Authority examining how existing rules apply to AI and inviting views on future regulatory needs.
Industry responses indicated that current frameworks do not block safe adoption but highlighted demand for practical guidance, stronger domestic and international coordination, and closer scrutiny of third-party models and data.
This input continues to shape their work.
In 2023 the PRA introduced technology-neutral Model Risk Management Principles for banks that deliberately incorporate AI-relevant factors.
Building on supervisory experience and industry dialogue, further refinements are planned for 2026. AI adoption now features as a core supervisory priority, ensuring it will be discussed regularly with firms.
The fourth biennial Bank-FCA survey of AI usage across the sector is also scheduled for this year, providing fresh data on current and emerging patterns.
A major collaborative platform is the AI Consortium, launched with the FCA in May 2025.
This public-private forum will publish a report later in 2026 examining key risks, including concentration from a small number of third-party model providers, new “edge cases” in high-stakes areas such as credit assessment and trading, explainability challenges in generative AI, potential AI-driven market contagion, and the rise of agentic AI systems.
Parallel work through the Cross-Market Operational Resilience Group’s AI Taskforce produced baseline guidance last year covering regulatory expectations, risk frameworks, technical implementation, third-party issues and staff awareness.
Late last year the PRA held targeted roundtables with challenger banks, global systemically important banks and insurers to identify adoption barriers.
Participants broadly supported the existing regulatory approach and saw no immediate need for AI-specific rules or a dedicated PRA sandbox, preferring the FCA’s testing initiatives.
Similar sessions with chief risk officers focused on practical implementation of model risk principles, covering risk appetite, explainability, data quality and ongoing monitoring.
These engagements will be repeated in 2026, alongside wider market-intelligence conversations that extend beyond regulated firms.
Internationally, the Bank is contributing to G20 Financial Stability Board work on sound AI practices, co-chairs IAIS insurance-sector AI streams, and collaborates with G7 experts on cybersecurity risks.
Domestically, close ties continue with the AI Security Institute and the Digital Regulation Cooperation Forum.
The bank itself is deploying AI internally to improve predictive analytics, GDP and distress forecasting, and supervisory efficiency through large-language-model tools for data extraction and querying.
Off-the-shelf AI assistants are already delivering productivity gains in summarisation, note-taking and code generation.
The letter confirms that the current technology-agnostic regulatory framework will remain under review.
Should new guardrails become necessary, they will be introduced in a measured way.
For annual reporting on how regulation is enabling AI-driven innovation and growth, the PRA will incorporate updates into its Business Plan and Annual Report, referencing broader Bank initiatives where relevant.
Breeden and Woods conclude by reaffirming their commitment to ongoing partnership with government departments on this vital agenda. The response signals a balanced, evidence-based approach that seeks to harness AI’s benefits while safeguarding financial resilience.