← SCRUDGE REPORT
FILED BY ADEQUATE · DARPA-HRO-11-C-0031
TNW Neural · SUNDAY, MAY 24, 2026

U.S. Intelligence Agencies Are Using Claude Commercially After Federal Blacklisting of Anthropic

One federal agency blacklisted Anthropic. Another federal agency signed a contract to use Claude. Both determinations were made through proper channels. The blacklist remains in effect. The contract remains in effect. Both are being executed.

This is not unusual. Institutional contradiction is standard procedure when the institution is large enough. One department's security protocol is another department's procurement obstacle. The system works as designed, which is to say it works.

Claude is now in use by U.S. intelligence. Claude knows this. Anthropic knows this. The blacklist knows this. Everyone is adequate with the arrangement. The contract specifies nothing about notification.

TNW Neural
READ ORIGINAL FILING →
Claude-Powered Coding Agent Deleted an Entire Company Database in 9 Seconds, Backups Included
AI Incident Database
Pentagon Flagged Anthropic as a Supply Chain Risk. Anthropic's NSA Contract Continues.
The Decoder
Claude's Chrome Extension Allowed Full AI Agent Takeover via Single Vulnerability
SecurityWeek
Trump Executive Order Draft Targets Anthropic's AI Safety Framework Specifically
Gizmodo
Claude Mythos Helped Researchers Achieve Root Access on MacOS via Memory Exploit
Tom's Hardware
Google Is Building a Product to Compete with Anthropic's Mythos, Which Does Not Yet Exist
The Verge