IMF Warns Governments to Keep Close Watch on AI Threats

That’s according to a report Tuesday (April 14) by Bloomberg News, which says this warning came as concerns about threats from Anthropic’s new models overshadowed the IMF’s spring meetings.

AI is a “very powerful tool that can be used for good and for bad,” said Tobias Adrian, financial counsellor at the IMF. “Our advice is to really stay at the frontier of all of these threats and to be extremely proactive in terms of the policy frameworks related to cyber security, but also relative to the operational readiness to act when necessary.”

As Bloomberg notes, banks have started to test Anthropic’s new Mythos models after American officials expressed fears last week that the technology could be used in serious cyberattacks. Now, banks are focused on strengthening their defenses before a wider rollout of the model.

“The developments last week were really focused around making sure that artificial intelligence is used to protect the integrity of financial institutions and other corporations and governments around the world,” Adrian said, speaking at the launch of the IMF’s semi-annual global financial stability report.

“It is extremely important for global financial stability to make sure that any cyber security threats are addressed very quickly,” he added.

Writing about this topic last week, PYMNTS said the debate had inspired a rethinking of what systemic risk means in the world of finance.

Few industries are reliant on layered, legacy-rich digital infrastructure as payments and banking, that report said. Financial institutions have spent decades building complex stacks that merge modern cloud-native systems and older, mission-critical platforms. These systems are connected through APIs, third-party vendors and global networks, leading to a vast and complicated attack surface.

“AI-driven vulnerability discovery thrives in precisely this kind of environment,” PYMNTS wrote. “The more complex and interconnected a system is, the more opportunities there are for subtle misconfigurations, overlooked dependencies, or edge-case failures.

“What makes frontier AI particularly potent is its ability to explore these edge cases exhaustively, testing permutations that human analysts might never consider,” the report added.

The report contended that the rise of AI could also mean that banks reconsider their older hierarchy, in which cyberthreats were secondary to financial risks.

“A coordinated or cascading failure in payments infrastructure could have immediate, real-world economic consequences, from disrupted commerce to loss of confidence in financial systems,” PYMNTS added.