Anthropic blocks 'malicious use' of AI that could develop biological weapons

Anthropic has identified and disrupted malicious activities using its AI models that could support the development of biological and conventional weapons, according to the company's latest threat intelligence report. The report reveals that Claude, one of Anthropic's AI models, was used in a Russia-linked cyber espionage campaign and by an Iranian propaganda institution between December 2025 and August 2026.

Additionally, the company detected misuse of its models for fake dating apps, hotel Wi-Fi scams, surveillance to identify dissidents, and the development of software for conventional weapons like firearms, missiles, drones, bombs, and targeting systems. Anthropic believes it has a responsibility to disclose such misuse and has shared intelligence with authorities and industry partners.

The report highlights five case studies of actors using Anthropic's models for biological weapons development, emphasizing that such misuse is one of the most serious risks of frontier AI models.