BOE's Bailey Talks AI Risks, Private Credit and Crypto

Andrew Bailey discusses AI cyber risks, the need for international cooperation on AI model access, and the interconnectedness of private credit with the banking system. He also touches on stablecoin regulation, tokenized deposits as a preferred path for digital payments, and the Bank of England's use of AI for data analysis and communication.

inferred

inferred

Bank of England 9.0
Central Bank
Andrew Bailey 7.0

BOE's Bailey Talks AI Risks, Private Credit and Crypto

dxy

Bailey notes the US sees its approach to stablecoins as a competitive advantage, but the rest of the world will take a different view. This suggests ongoing divergence but no clear directional signal for the dollar index in the near term.

metals

No direct mention of metals. The discussion is centered on financial technology and credit markets, not commodities.

ndx

Bailey is optimistic about AI development and its benefits, but emphasizes the need for careful risk management and regulation. This suggests a cautiously positive long-term view for the tech-heavy NDX.

rut

Bailey's discussion of private credit and its interconnection with banks is relevant to smaller, domestically-focused companies in the Russell 2000. The cautious tone on financial stability suggests a neutral to slightly cautious outlook.

wti

No direct mention of oil or energy markets. The focus on financial stability and AI does not provide a clear directional signal for WTI.

yields

Bailey's focus on financial stability, interconnectedness of private credit, and the need for international rules suggests a cautious, status-quo approach to monetary policy and yields, with no strong directional bias.

Francine Lacqua

Asks about UK banks' access to the AI model 'Mythos' (likely a reference to a specific model, possibly a misnomer for 'Claude' by Anthropic) and the associated cyber security risks.
Andrew Bailey

Confirms UK banks do not yet have access to Mythos, but are optimistic about access to other models. Notes the delay is caught up in US administration processes.
Francine Lacqua

Raises the issue of open-source AI models being a vulnerability, as they cannot be turned off if used by malicious actors.
Andrew Bailey

Agrees that open-source models present a greater vulnerability than closed ones, based on work done by Anthropic.
Francine Lacqua

Questions the risk of US dominance in AI and UK banks having to 'go begging' for access.
Andrew Bailey

Acknowledges the development is very big and important, but the spillover cyber risks are so large that a single national approach won't work; an international approach is required.
Francine Lacqua

Asks about the risks of private credit and its interconnection with traditional banks, and whether the US is receptive to tougher reporting requirements.
Andrew Bailey

Views private credit as a good thing that supports investment demand, but warns it is not an island and is connected to the banking system. The Bank of England is doing a second system-wide exploratory scenario to understand these interconnections and points of fragility.
Francine Lacqua

Asks about the UK's approach to stablecoin regulation, specifically whether the idea of capping individual holdings is being scrapped.
Andrew Bailey

Confirms the UK will come out with new proposals before the summer, and that total issuance limits are an option. Emphasizes the need for international 'rules of the game' for cross-border stablecoin use.
Francine Lacqua

Points out the lack of US appetite for such rules, as they see it as a competitive advantage.
Andrew Bailey

Believes the rest of the world will take a different view on stablecoin regulation.
Francine Lacqua

Asks about the Bank of England's digital pound project, given the US has banned a digital dollar.
Andrew Bailey

Expresses skepticism about the need for a new form of central bank money. Suggests the natural starting point is to explore tokenized deposits, which is already happening domestically and internationally (e.g., Project Agora).
Francine Lacqua

Asks how the Bank of England is using AI tools.
Andrew Bailey

Confirms the Bank of England uses AI for data analysis, coding, and modeling. They also use language models to test how MPC minutes might be received by the public.

SignalTube

markets at a glance