A powerful AI model may now be enough to help attackers threaten parts of the global financial system, the International Monetary Fund (IMF) warned recently.

In a new assessment, the IMF said artificial intelligence is making cyberattacks faster, cheaper and easier to scale, posing a growing threat to banks, payment networks and shared financial infrastructure worldwide.

The organization warned that advanced AI systems can dramatically reduce the time needed to identify and exploit software vulnerabilities, giving attackers the ability to strike multiple institutions before defences have a chance to respond.

That risk is amplified by the financial sector’s growing dependence on shared cloud providers, software platforms and interconnected digital systems. A single breach, the IMF cautioned, is no longer a local problem. It can ripple across institutions and shake broader markets.

The warning comes amid rising concern over the cyber capabilities of frontier AI systems. The IMF pointed to Anthropic’s Claude Mythos Preview, saying the model could “find and exploit vulnerabilities in every major operating system and web browser,” a capability the IMF said requires no expert to deploy.

The IMF said emerging economies could face greater exposure because cybersecurity protections and recovery resources are often weaker.

The IMF also stressed that AI is becoming a key defensive tool. Financial institutions are increasingly using AI systems to detect fraud, monitor threats and identify security gaps before attackers exploit them.

Still, the IMF warned that technology alone will not be enough. It called for stronger cyber resilience standards, systemic stress testing and greater coordination between governments and private institutions.

Also Read: India’s Finance Minister Warns Banks on Claude Mythos AI Model Cyber Risks