US Treasury Secretary Praises Anthropic’s Mythos as Key to AI Edge Over China

US Treasury Secretary Scott Bessent called Anthropic’s Mythos a breakthrough in keeping the United States ahead of China in artificial intelligence. Speaking at a Wall Street Journal event in Washington, Bessent said American AI remains three to six months ahead of China’s capabilities.

Bessent highlighted Mythos, a model developed by Anthropic that is designed to detect vulnerabilities in software and computer systems. The model is being released only to a small group of selected organizations. He described it as a major leap in learning performance, comparing its progress to exponential growth.

His remarks came shortly after Treasury and Federal Reserve officials held an emergency meeting with Wall Street banks over concerns that Mythos could increase cybersecurity risks. Despite tensions between Anthropic and federal agencies, including the Pentagon labeling the company a supply chain threat, Bessent endorsed Anthropic as a leading force in the AI sector.

Anthropic, founded in 2021 by former OpenAI employees, has focused on enterprise AI systems and is investing heavily in U.S. data center infrastructure, with a planned $50 billion expansion.