Bank of England, BIS warn AI risks could destabilize global financial markets

Central bankers and economists at the European Central Bank’s annual forum in Sintra said artificial intelligence could become a source of financial instability through multiple channels, from asset-price bubbles and opaque lending decisions to cyber threats and labor-market disruption. The discussion broadened earlier warnings from the Bank of England and the Bank for International Settlements, with speakers arguing that both an AI boom that overdelivers and one that falls short could create risks for the financial system. Torsten Slok of Apollo Global Management said AI could affect financial stability whether it exceeds expectations or disappoints, while Federal Reserve Chairman Kevin Warsh described the technology as one of the biggest economic turning points of their lifetime. In markets, University of Pennsylvania professor Itay Goldstein warned that more advanced algorithms could coordinate on manipulative price paths, potentially inflating bubbles and accelerating crashes. The BIS said in a report that the scale and speed of the current AI investment boom, and expectations of large productivity gains, resemble earlier episodes such as railway mania, the 1920s and the dotcom boom, highlighting near-term downside risks. Officials also said AI could improve credit analysis and expand lending, but make supervision harder because agentic loan decisions can operate as black boxes. Tobias Adrian, a senior IMF official, said the lack of explainability is a key supervisory challenge. Sarah Breeden, a Bank of England Deputy Governor, said resilience tools analogous to deposit insurance may be needed in cyber disruptions so one institution can maintain another’s basic functions. Speakers also warned that if AI meaningfully displaces workers, falling incomes could weaken the economy, while a failure of heavy AI investment to generate expected returns could also produce a destabilizing unwind.