Central bankers warn agentic AI poses systemic financial risk
In brief
- Nikhil Rathi (FCA) warns traditional regulation cycles cannot keep pace with AI development
- Sarah Breeden (Bank of England) calls for circuit-breaker guardrails to limit agentic AI volatility
- Christine Lagarde (ECB) identifies AI as a more serious financial risk than cybersecurity
Regulators warn on AI speed and systemic risk
Nikhil Rathi, CEO of the UK's Financial Conduct Authority, told CNBC that traditional regulation cycles do not work in an era of fast-moving AI development. Technology moves incredibly fast, and regulators need new approaches to keep up with innovations in the AI market. Rathi emphasized the need for collaboration between the financial industry and watchdogs to manage emerging risks.
Bank of England deputy governor Sarah Breeden warned that agentic AI could amplify volatility during periods of market stress. She questioned whether guardrails analogous to circuit breakers or kill switches are needed to limit or stop trading if faulty AI models cause market meltdown.
Breeden also flagged rising debt financing tied to AI assets. The financial stability consequences of a sharp pullback in AI-related asset prices could be severe, she said, particularly if central banks tighten policy to contain inflation.
ECB and BIS escalate warnings
European Central Bank President Christine Lagarde warned that AI technology poses a major risk to financial systems. She emphasized that AI models present a more serious threat than cybersecurity because they are advancing very quickly and defense mechanisms have yet to be found.
The Bank for International Settlements warned that AI exuberance could have major financial consequences. The BIS cautioned that tightening central bank policy could trigger a sharp pullback in AI asset prices and disruptive macro-financial feedback loops.
Structural vulnerabilities emerge
Tobias Adrian, Director of the IMF's Monetary and Capital Markets Department, identified a potential maturity mismatch between the duration of physical AI assets and the duration of debt financing them. This structural imbalance could amplify losses if asset prices decline.
US companies are leading in AI investment and frontier model development. Europe's financial system provides fewer capital channels into AI compared to US equity markets, creating divergent exposure to AI-driven volatility across regions.
Regulators and central bankers have called for guardrails to protect the financial system from rapid advances in agentic artificial intelligence. The consensus is clear: traditional oversight cannot keep pace with AI's speed and complexity.
