News
AI Cyber Risks · Cybersecurity Investment · National Security · Tech Regulation
The White House, in collaboration with major tech firms including Microsoft, Alphabet, Amazon, and leading banks like Goldman Sachs and Bank of America, is urgently addressing escalating cyber risks posed by advanced AI models, prompting limited access to Anthropic's Mythos system due to its bug-finding capabilities.
Officials treat these risks as a near-term concern, focusing on potential cyber threats to critical U.S. systems. National Cyber Director Sean Cairncross leads efforts to identify weak points, with Kevin Hassett, director of the National Economic Council, stating there is "definitely a sense of urgency." Anthropic's Mythos model, highly effective at finding and exploiting software bugs, is not publicly released; instead, Anthropic provides limited access to approximately 50 key infrastructure firms, including Apple, Amazon, and Alphabet, to help them fix issues.
OpenAI and other developers prepare to launch new tools, raising concerns about bad actors using these systems for large-scale, rapid cyberattacks. The White House coordinates with the Office of Science and Technology Policy and the National Security Council, engaging private firms.
Senior officials held calls with executives from OpenAI, Microsoft, Alphabet, Anthropic, CrowdStrike, and Palo Alto Networks. Treasury Secretary Scott Bessent, Vice President JD Vance, and Federal Reserve Chair Jerome Powell met with major bank leaders from Goldman Sachs, Bank of America, and Wells Fargo to discuss responses to AI-linked cyberattacks.
Policy tensions exist, with the Trump administration limiting some Anthropic models in federal agencies. AI is now a key issue for cyber safety, financial stability, and national security in the U.S.
