Anthropic blocks AI use in bioweapons, Russian cyberespionage; reports Chinese infiltration attempts

Anthropic's threat intelligence report documents blocking Claude models from aiding biological weapons development and a suspected Russian cyberespionage campaign against Ukraine, while also accusing Chinese competitors of attempting to extract Claude's capabilities. This marks a notable escalation in AI-specific security threats, highlighting the dual-use risks of advanced AI and state-linked cyber operations targeting AI firms.