Anthropic report: AI used in cyberattacks, spying, bombs and pathogens
Anthropic, an AI company, just shared in a lengthy report that its AI technology has been used for some pretty serious stuff, like cyberattacks, spying, and even helping create bombs and dangerous pathogens.
The report was released on September 11, 2026 and highlights how both criminals and state-backed groups have attempted to use the technology for harmful purposes.
Jacob Coxon quits, Anthropic urges safeguards
Just two days before the report came out, former employee Jacob Coxon quit, accusing Anthropic of chasing advanced AI without enough safety checks.
The company also found researchers bypassing security to study viruses, which could help with vaccines but also risks creating biological weapons.
Anthropic has banned those accounts but hasn't said who was involved.
With cases like Russian espionage and Chinese surveillance popping up in the report, Anthropic is calling for stronger industry teamwork and government action to keep things safe as AI keeps evolving.