The Bank of England will review its guidelines for working with artificial intelligence in the financial sector. This was announced by Deputy Governor Sarah Breeden at the European Central Bank Forum in Sintra.
According to her, the current technology-neutral regulations were not designed for systems capable of operating without human intervention. The main risks of actively implementing AI agents include cyberattacks, payment disruptions, and synchronized market errors.
“The obvious acceleration in AI capabilities, even compared to six months ago, presents a dual challenge: not only ensuring responsible implementation and risk management but also recognizing that AI must transform central banking methods. We need to seriously consider both aspects,” Breeden said.
Among the possible measures, she mentioned stricter requirements for restoring key systems, fully isolated backup capacities, and the ability to quickly restructure compromised infrastructure. For markets, the regulator is also exploring safeguards similar to emergency trading halts in case erroneous AI models trigger large-scale disruptions. Breeden also highlighted the risk of herd behavior, where agents react to the same triggers and amplify volatility.
On May 15, the Bank of England, Financial Conduct Authority, and HM Treasury issued a joint statement on the cyber risks of advanced AI models for the financial sector. On June 22, a similar assessment was made by Five Eyes, warning that the window between vulnerability detection and exploitation is shrinking to months.
On June 10, the Financial Stability Board published a consultation document with 12 practices for the responsible adoption of AI by financial organizations. It does not establish a mandatory international standard but offers a framework for corporate governance and risk control at all stages of development and implementation.
According to data from the Cambridge Centre for Alternative Finance, 81% of surveyed industry participants are already implementing AI at some level, and 52% are at the pilot stage or more advanced deployment of agent AI. Currently, the main applications are internal processes: automation, data visualization, and software development. Front-office implementation remains limited.
In a response to a parliamentary committee on April 1, the Bank of England indicated that generative and agent AI are not yet used in the financial system in a way that creates systemic risk. However, the regulator emphasized that the technology-neutral approach will be reviewed.
In May 2026, the International Monetary Fund urged that AI cyber risks be considered a matter of financial stability, not just IT security.
The discussion is also ongoing in other institutions, including the Bundesbank and the Bank for International Settlements.
