Federal Reserve Governor Barr discusses several potential effects of artificial intelligence on financial stability
Some Potential Financial Stability Implications of Artificial Intelligence
Federal Reserve Governor Barr gave a speech systematically assessing the potential effects of artificial intelligence on financial stability. Citing the Federal Reserve's latest Financial Stability Report, he said the on-balance-sheet debt of the seven largest AI companies is expected to exceed $800 billion by the end of the year, and the potential maximum notional value of their off-balance-sheet liabilities already exceeds $3 trillion; hyperscale cloud providers and GPU vendors have also made more than $100 billion in circular equity investments in LLM labs and cloud computing startups. Barr warned that trading reliant on similar AI models could trigger herd effects, and that some reinforcement learning algorithms may spontaneously collude or spoof to manipulate markets; AI attackers are already able to exploit vulnerabilities previously considered difficult to overcome. On consumer finance, large banks have found AI agents attempting to operate retail customer accounts but are unable to verify their authenticity and authorization; he warned that agents could accelerate deposit outflows during stress periods and cause a run to unfold within minutes, and called for guardrails such as strict identity verification, authorization verification and limiting agent capabilities.