Bank of England Governor Warns AI Threats Need Global Teamwork, Not Just US Action
Andrew Bailey says no single country can handle the risks of powerful AI models alone, days after the Trump administration blocked foreigners from accessing one of the most advanced AI tools on the market.
Key points
- Bank of England Governor Andrew Bailey called for international cooperation to manage AI risks, in comments reported by The Guardian AI.
- Weeks before Bailey spoke, US President Donald Trump temporarily banned non-US users from accessing Claude Mythos, a powerful AI model built by the company Anthropic.
- Bailey warned directly that the United States would not be able to meet its AI ambitions without working with other countries.
- The intervention marks one of the most senior interventions on AI policy from a major central bank governor.
Andrew Bailey, the governor of the Bank of England, Britain's central bank and the institution that sets interest rates and watches over the financial system, has delivered a pointed warning to Washington. No country, he said, including the United States, can manage the dangers of modern AI on its own.
The timing matters. His comments follow a decision by President Donald Trump to temporarily bar foreigners from using Claude Mythos, a highly capable AI model, meaning software that can hold conversations, write code, and reason through complex problems, made by the American company Anthropic. That move struck many observers as a signal that the US sees advanced AI as a national asset to be guarded, not shared.
Bailey pushed back on that logic.
His argument is straightforward: AI systems cross borders the moment someone opens a browser. Threats they create, whether financial fraud, misinformation, or destabilising markets, do not stop at customs. A purely national response, he suggested, leaves gaps that bad actors will find quickly.
What does this mean for ordinary people?
For most people, a speech by a central bank governor sounds remote from daily life. It is not. If AI-driven financial fraud scales up faster than regulators can respond, ordinary bank customers are the ones who absorb the losses or face the disruption. International rules, agreed between governments, are the main tool that stops that from happening.
Bailey's call echoes what a number of European regulators have said for months: that setting safety standards country by country creates a patchwork that protects nobody properly. The EU passed its AI Act, a sweeping law that sorts AI systems by how much risk they carry and sets rules accordingly, earlier this year. The US has so far preferred voluntary industry commitments over hard law.
The gap between those two approaches is exactly what Bailey appears to be worried about.
Whether Washington listens is another question. The Trump administration has framed AI leadership as a matter of national competitiveness. Sharing oversight frameworks with international partners does not sit easily with that framing.
For now, Bailey's intervention adds serious institutional weight to the case for a coordinated global approach. Whether that pressure translates into actual policy talks is the question that will shape how AI develops, and how safely, over the next few years.
