N.S.A. Lost Access to Powerful A.I. Model Amid Anthropic Dispute

Key Points:

• The National Security Agency (NSA) has lost access to Anthropic's advanced A.I. models, Mythos 5 and Fable 5, due to export controls imposed by the Trump administration over national security concerns.
• These A.I. tools had impressed NSA cybersecurity analysts with their ability to identify software vulnerabilities quickly, reportedly breaching most classified systems within hours during controlled tests.
• The export restrictions forced Anthropic to halt the release of its most powerful models, disrupting the NSA's ongoing cybersecurity efforts.
• The incident highlights the growing reliance of U.S. intelligence agencies on cutting-edge A.I. technology for cybersecurity, even amid regulatory and political conflicts with domestic developers.
• Senator Mark Warner emphasized the significance of these tools during a congressional hearing, underscoring the tension between national security priorities and technological innovation.