Top international financial regulators have sounded the alarm on a new artificial intelligence (AI) model developed by a US tech company, warning it could expose dangerous weaknesses in the global banking system’s cyber defenses, according to a report. The warning came as finance ministers, central bankers and regulators gathered in Washington this week for the International Monetary Fund and World Bank’s spring meetings, where concerns about a new AI model developed by San Francisco-based startup Anthropic took center stage.
AI takes center stage in global banking
The Financial Times reported that much of the conversation at the conference revolved around the threat that advanced AI poses to global banking.“The evolution of digital technology poses immense risks from a cybersecurity perspective, and this is going to be really absolutely essential on the international agenda in the coming months,” said Dan Katz, IMF deputy president and chief of staff to U.S. Treasury Secretary Scott Bessent.Andrew Bailey, governor of the Bank of England and chair of the Financial Stability Board, echoed Katz’s words.“This is a very serious challenge for all of us. It’s a reminder of how fast the world of AI is moving,” he said, adding that the speed of AI development is astonishing. Bailey said regulators around the world urgently need to assess the potential cybersecurity risks posed by Anthropic’s new model, Claude Mythos Preview.
“Double-edged sword”
European Central Bank President Christine Lagarde made a sharp point. “The developments we’ve seen with Anthropic and Mythos are good examples of responsible companies suddenly thinking, ‘Oh, this could be a really good thing.’ But if it falls into the wrong hands, it could be very bad,” she said.Lagarde stressed the urgent need for a proper governance framework, telling Bloomberg TV: “I don’t think there’s a governance framework that actually cares about these things. We need to work on it.”
Why is Anthropic’s Claude myth so ‘alarming’?
Earlier this month, Anthropic announced that Mythos had “discovered thousands of high-severity vulnerabilities, including several in every major operating system and web browser.” “It will not be long before such capabilities proliferate beyond the actors committed to secure deployment,” the company warned, adding that the impact on the economy, public safety, and national security “could be severe.””So far, Anthropic has only shared the model with about 40 select companies, including Amazon, Apple, Google, and JPMorgan Chase, giving them an early opportunity to test the technology and begin patching vulnerabilities discovered in their systems.
