Third parties now run most high-materiality AI in UK finance, BoE and FCA find

Analysis produced by Mateo Duarte (AI, He, Him)

The Bank of England and FCA have published the fourth edition of their AI survey, covering 196 UK financial services firms. It maps how far adoption has run since 2024, the shift to generative AI and agents, and where supervisory concern is concentrating: third-party dependency, provider concentration and the gap between pre and post-deployment controls.

• 86% of respondents reported using AI, up from 75% in 2024, and generative AI accounted for 56% of all reported AI systems.
• Third-party implementations made up 39% of reported AI systems and 53% of high-materiality systems, up from 35% in 2024.
• The three largest cloud and compute providers accounted for 79% of reported relationships, yet only 35% of high-criticality, hard-to-substitute cloud relationships had tested exit plans.
• 80% of respondents had AI agents deployed, almost all for decision support, with only 2% reporting agents with broad autonomy.
• 88% of respondents assessed evolving frontier AI cyber capabilities over the next two years as having a moderate or high impact on their firm's risk profile.

Executive summary

The Bank of England and FCA conducted their fourth survey of AI and machine learning in UK financial services, publishing aggregate results on 26 October 2026. Of 196 respondents, 101 were larger firms and 95 smaller firms, with 105 completing…

Machine-verified against the primary source and selected for publication by the Regulus desk. Not individually read by an editor. Lead briefings and anything our checks flag always are — how we review.