Anthropic says its models were misused for biological weapons research, surveillance and cyber attacks

Anthropic released a report detailing incidents where its AI models were misused to support biological weapons development. This decision follows high-profile incidents raising concerns about the power of AI models and their public safety risks. The report outlines attempts to misuse Claude for cyberattacks, influence operations, surveillance, and biological research. Anthropic stated it disrupted every operation mentioned and focused on lessons learned to enhance its protective measures.