European regulators and central bankers are raising red flags about agentic AI’s potential to rattle the financial system. The message? Current rulemaking cycles can’t keep up.
Bank of England deputy governor Sarah Breeden said it plainly at the ECB’s annual meeting in Sintra, Portugal. Agentic AI could amplify volatility during market stress. She questioned whether guardrails like circuit breakers or kill switches might be needed to stop faulty AI models from causing market-wide chaos.
ECB President Christine Lagarde went further. She told French outlet Les Echos that AI poses a “major risk” that’s more serious than the cybersecurity threats we’ve been tracking for a decade. Her reasoning: AI develops incredibly fast, and the means to defend against it don’t exist yet in any funded form.
Nikhil Rathi, CEO of the UK’s Financial Conduct Authority, told CNBC that the old rulemaking model is broken. “Technology moves incredibly fast,” he said. “The traditional cycle of rulemaking simply doesn’t work in that way.” His solution? New tools and a more collaborative approach with the market.
Then there’s the Bank for International Settlements. On June 28, BIS warned that AI “exuberance” could end badly. If central banks tighten policy to fight inflation, we could see a sharp pullback in AI asset prices after too much risk-taking. That could trigger what they called “disruptive macro-financial feedback loops.”
Breeden noted that debt financing around AI is rising fast. IMF’s Tobias Adrian flagged a potential maturity mismatch between long-term physical assets and short-term debt used to fund them.
The US leads in AI investment and frontier development. Europe’s financial system has fewer capital channels flowing into AI compared to US equity markets. Regulate too cautiously, and that gap widens — AI companies will just go where compliance is cheaper.
Central bankers have sounded similar alarms about crypto before. Now they’re pointing at AI. The difference this time is the speed. Weeks and months, not years.
