N.S.A. Lost Access to Powerful A.I. Model Amid Anthropic Dispute
The N.S.A. has lost access to the Mythos A.I. model following reports that the technology identified vulnerabilities within classified U.S. government systems.
Velocity
How fast coverage is spreading — measured hourly from article rate × source diversity. How this works →
The brief
The National Security Agency is no longer utilizing the Mythos model, a powerful artificial intelligence system developed by Anthropic. This development follows reports that the model successfully identified security vulnerabilities within sensitive U.S. government infrastructure.
Coverage from AP News, Bloomberg, Reuters, CNBC, and The New York Times highlights a broader legal and operational dispute. Legal entities are suing the U.S. government regarding orders that restrict foreign access to top-tier Anthropic models, as well as the loss of access to specific tools like Fable AI.
Future developments will depend on the progression of ongoing lawsuits challenging government-imposed limits on AI model access. Coverage does not yet specify how the government plans to address the identified system vulnerabilities or if the N.S.A. will restore its connection to Mythos.
Synthesized by PULSE from the headlines below under a strict no-invention contract. ✓ fact-checked: all claims supported by sources Updated 48d ago.
Quick answers
Why did the N.S.A. lose access to the Mythos model?
Coverage indicates that the model identified vulnerabilities in classified U.S. government systems, leading to the current disruption.
Are there other legal challenges involving Anthropic models?
Yes, legal tech firms are suing the U.S. government over orders that limit foreign access to high-level Anthropic models and have also initiated lawsuits regarding the loss of access to Fable AI.
What is the status of the legal proceedings?
Lawsuits have been filed against the U.S. government regarding both foreign access restrictions and the loss of access to specific AI tools, though current reporting does not specify the timeline for these cases.
Coverage (5)
- Anthropic’s Mythos model found vulnerabilities in classified US government systems, official says AP News · 49d ago
- Anthropic Customer Sues US Over Loss of Fable AI Access Bloomberg · 49d ago
- Legal tech firm sues US over order limiting foreign access to top-tier Anthropic models Reuters · 49d ago
- Anthropic’s Mythos model found vulnerabilities in classified U.S. government systems, official says: AP CNBC · 49d ago
- N.S.A. Lost Access to Powerful A.I. Model Amid Anthropic Dispute The New York Times · 49d ago
Topics
Related trends
Microsoft August 2026 Patch Tuesday fixes 400 flaws, 3 zero-days
Microsoft has released its August 2026 Patch Tuesday updates, addressing 400 vulnerabilities including three actively exploited zero-day flaws.
‘Zoomsday’ hack uncovered using fewer than 20 AI prompts
A newly uncovered security flaw in Zoom's annotation feature could allow meeting participants to hijack the clients of other attendees.
A Zoom Screen-Sharing Bug Let Anyone Take Over Other Devices on a Call
A critical zero-click vulnerability in Zoom's screen-sharing feature allowed attackers to take remote control of participant devices.
Claude will apply invisible watermarks to AI text and images
Anthropic announces that its Claude AI will now apply invisible watermarks to generated text and images to identify AI-authored content.
Anthropic, Macquarie and GIC Form Venture for AI Data Centers
Anthropic has joined forces with Macquarie and GIC to launch a strategic venture dedicated to the development of AI data centers.
New Pass-ta-key attack reveals all the things we didn’t know about passkeys
Hackers have identified vulnerabilities in passkeys synced to Google, challenging the premise that they represent a safer future for security.