The UK has deliberately chosen not to write an AI rulebook for finance. The Bank of England, PRA, and FCA regulate AI through existing technology-agnostic frameworks — most recently reaffirmed in the BoE/PRA's April 2026 response to Parliament's inquiry on AI in financial services. Instead of rules, the UK runs the deepest supervisory monitoring program anywhere: periodic AI surveys (2019, 2022, 2024, and a new one covering foundation models and agentic AI), sandbox initiatives, and AI as a named 2026 supervisory priority.
| Full name | Bank of England, Prudential Regulation Authority, and Financial Conduct Authority |
| Role | Prudential and conduct regulators |
| Force on banks | Supervisory guidance |
| Applies to | UK banks, building societies, investment firms, and insurers |
| Key document | BoE/PRA response to the Treasury Select Committee on AI (Apr 2026) |
| Latest move | Apr 2026: tech-agnostic approach reaffirmed; AI named a 2026 PRA supervisory priority; new AI survey covering foundation and agentic AI |
The UK approach was set by DP5/22 (2022) and the FS2/23 feedback statement (2023): existing frameworks — the Senior Managers regime, model risk management (SS1/23), operational resilience, and Consumer Duty — already cover AI, so the regulators supervise outcomes rather than technology. The government reinforced this in January 2026 with strategic letters directing 19 regulators, including the FCA and Bank of England, to publish plans for enabling safe AI innovation and report annually.
On April 1, 2026 the BoE and PRA responded to the Treasury Select Committee's inquiry on AI, reiterating the technology-agnostic stance while keeping 'whether further guardrails are needed' under review. AI adoption is a PRA supervisory priority for 2026, meaning firms face detailed supervisory dialogue on their AI governance even without new rules. The FCA, meanwhile, has invested in AI sandbox and live-testing initiatives to observe deployments directly.
| Date | Type | Document / event |
|---|---|---|
| Oct 11, 2022 | Consultation | DP5/22: Artificial Intelligence and Machine Learning. Joint BoE/PRA/FCA discussion paper asking whether existing regulation is sufficient for AI in financial services. |
| Oct 26, 2023 | Report | FS2/23 feedback statement. Concludes existing frameworks largely suffice; regulators will remain technology-agnostic and monitor via surveys. |
| Nov 21, 2024 | Survey | Third BoE/FCA AI survey in UK financial services. Finds 75% of firms already using AI; foundation models a fast-growing share; third-party exposure concentrated. |
| Jan 1, 2026 | Framework | Government strategic letters on AI to 19 regulators. DSIT and DBT direct regulators including the FCA and Bank of England to publish plans by May 2026 for enabling safe AI-powered innovation and to report annually. |
| Apr 1, 2026 | Report | BoE/PRA response to the Treasury Select Committee inquiry on AI. Reaffirms the technology-agnostic approach, sets out a roadmap for responsible AI adoption, and keeps the need for further guardrails under review. |
| Jun 1, 2026 | Survey | New AI survey covering foundation models and agentic AI. Fourth-generation voluntary survey of AI/ML use in UK financial services, extended to foundation models, generative AI, and agentic AI. |
- Output of the new AI survey — the first regulator dataset anywhere on agentic AI in banks
- Whether rising political scrutiny pushes the PRA/FCA from monitoring toward formal guardrails
- FCA AI live-testing and sandbox findings feeding into supervisory expectations
Does the UK have an AI law for banks?
No. The UK has explicitly chosen not to create AI-specific rules for financial services. The Bank of England, PRA, and FCA apply existing technology-agnostic frameworks — senior manager accountability, model risk management (SS1/23), operational resilience, and the Consumer Duty — to AI, a position most recently reaffirmed in April 2026.
How do UK regulators monitor AI in banks?
Primarily through recurring joint BoE/FCA surveys (2019, 2022, 2024, and a 2026 edition extended to foundation models and agentic AI), supervisory dialogue under the PRA's 2026 priorities, and FCA sandbox/live-testing programs that observe real deployments.
How does the UK approach differ from the EU AI Act?
The EU imposes binding, AI-specific obligations with credit scoring classified as high-risk; the UK regulates the same activities through existing outcome-based rules with no AI-specific statute. UK banks operating in the EU must still comply with the AI Act for their EU business.
Follow every move these regulators make
6 curated AI stories for banking executives · Every morning · Free
Subscribe to BankingNewsAI →