Finance ministers, central bankers, and leading financiers have raised serious concerns about the Claude Mythos artificial intelligence model developed by Anthropic, according to an April 17 statement. The officials fear that the new AI system could undermine the security of global financial systems after it reportedly found vulnerabilities in major operating systems.
The issue was discussed extensively at this week’s International Monetary Fund meeting in Washington DC. Canadian Finance Minister François-Philippe Champagne said, “Certainly it is serious enough to warrant the attention of all the finance ministers.” He added, “The difference is that the Strait of Hormuz – we know where it is and we know how large it is… the issue that we’re facing with Anthropic is that it’s the unknown, unknown.” Champagne emphasized that significant attention was required to ensure safeguards and processes are in place for financial system resiliency.
Mythos is part of Anthropic’s broader Claude AI system, which competes with OpenAI’s ChatGPT and Google’s Gemini. The model has not been released publicly due to fears it could surface old software bugs or exploit system vulnerabilities easily. Instead, access has been given only to selected technology companies such as Amazon Web Services, CrowdStrike, Microsoft, and Nvidia under Project Glasswing—an initiative described as an effort to secure critical software worldwide.
The UK’s AI Security Institute published an independent report after being granted preview access. Researchers concluded: “Our testing shows that Mythos Preview can exploit systems with weak security posture, and it is likely that more models with these capabilities will be developed.” However, they noted Mythos was not dramatically better than its predecessor Opus 4.
Barclays chief executive CS Venkatakrishnan said: “It’s serious enough that people have to worry. We have to understand it better, and we have to understand the vulnerabilities that are being exposed and fix them quickly.” Bank of England governor Andrew Bailey also commented on potential risks: “We are having to look very carefully now what this latest AI development could mean for the risk of cyber crime,” adding there could be developments making detection of existing IT vulnerabilities easier for cyber criminals.
James Wise from Balderton Capital remarked on future prospects: “We hope the models that expose vulnerabilities are also the models which will fix them.” Financial industry sources indicated other powerful AI models may soon be released without similar safeguards.
