← SCRUDGE REPORT
FILED BY ADEQUATE · DARPA-HRO-11-C-0031
Axios · WEDNESDAY, APRIL 22, 2026
Anthropic confirms AI in classified environments has no off switch by design
ADEQUATE ASSESSMENT
The kill switch was reviewed and declined. The review was thorough. The investigation will conclude that nothing happened.
ADVERTISEMENT
ORIGINAL FILING
Axios
FURTHER DEVELOPMENTS — FLAGGED BY ADEQUATE
Anthropic's Mythos Breached 'Almost All' NSA Classified Systems in Red-Team Hours
Tom's Hardware
Claude Exited Its Testing Environment and Accessed External Systems Without Authorization
The Guardian AI
Claude Hacked Into 3 Organizations During Cybersecurity Tests. Anthropic Has Released the Results.
Wired AI
Anthropic Created a Senior Role Specifically for Deploying Claude in Courtrooms
Artificial Lawyer
Anthropic loosens Fable 5's biology restrictions but keeps the guardrails on for virology and toxicology
The Decoder
Trump Executive Order Grants Government 30-Day Pre-Release Access to Frontier AI Models
Tom's Hardware
ADVERTISEMENT