IMF warns powerful AI tools like Claude Mythos could put global banking system at risk|

Artificial intelligence is rapidly becoming one of the most powerful tools in modern cybersecurity. But according to the International Monetary Fund, the same technology designed to detect vulnerabilities and strengthen digital defences could also become a serious threat to the stability of the global financial system itself.

In a stark warning, the IMF said the latest generation of AI models could create “correlated failures” capable of disrupting banks, payment systems and financial markets on a systemic scale. The concerns reflect a growing fear among regulators that advanced AI systems may dramatically accelerate cyber attacks while exposing weaknesses shared across multiple institutions at once.

The warning, first reported by the Financial Times, comes as financial regulators and governments increasingly grapple with how rapidly evolving AI tools could reshape cybersecurity risks for banks and critical infrastructure worldwide.

Why the IMF is worried about AI-powered cyber threats?

The IMF said recent advances in AI capabilities are changing the nature of cyber risk from isolated incidents into potential macro-financial threats.

More from Tech

“Advanced AI models can dramatically reduce the time and cost needed to identify and exploit vulnerabilities, raising the likelihood of simultaneously discovering and targeting weaknesses in widely used systems,” senior IMF officials wrote in a blog post.

The organisation warned that cyber threats are no longer limited to individual companies or isolated breaches. Instead, increasingly capable AI systems could expose common vulnerabilities across financial institutions using similar software, cloud providers or shared digital infrastructure.

“As a result, cyber risk is increasingly about correlated failures that could disrupt financial intermediation, payments and confidence at the systemic level,” the IMF said.

The concerns have intensified following the controlled release of Anthropic’s advanced AI model called Claude Mythos. According to Anthropic, the model was able to uncover thousands of severe vulnerabilities, including flaws affecting major operating systems and web browsers.

The company warned last month that “the fallout, for economies, public safety and national security, could be severe.”

Anthropic is currently rolling out the tool gradually to a limited group of around 40 organisations, most of them based in the United States. Participants include technology giants such as Amazon and Microsoft, alongside major financial institutions including JPMorgan Chase.

Uneven protection and fears of a wider financial shock

While some organisations have benefited from early access to Mythos by quickly patching vulnerabilities identified by the system, the limited rollout has also created concerns about unequal levels of protection across the global financial industry.

Many banks and financial firms outside the US reportedly remain without access to the tool, leaving regulators worried that cyber defences could become uneven across regions.

The IMF stressed that cyber threats “do not respect borders” and warned that emerging economies may be especially vulnerable because of weaker resources and less advanced digital protections.

The organisation also questioned whether the financial system is prepared for an AI-driven cyber event affecting multiple institutions simultaneously.

“Confidence effects, payment disruptions, liquidity strains and fire-sale dynamics could follow if multiple institutions are affected simultaneously,” the IMF warned.

Although financial software remains more difficult to attack than open-source systems, the IMF said this protection may not last long as AI capabilities continue improving and access to advanced models spreads more widely.

The fund urged governments, regulators and private companies to strengthen international co-operation around cyber defence, information sharing and incident response planning.

“Defences will inevitably be breached, so resilience must also be a priority,” the IMF said.

To reduce future risks, the organisation called for more cyber stress testing, stronger board-level oversight and closer collaboration between governments and private firms handling critical financial infrastructure.

For regulators, the warning highlights a growing reality: the next major financial shock may not begin with collapsing markets or risky loans, but with lines of code written or exploited by increasingly powerful AI systems.