Anthropic says it has detected and is warning about potential misuse of its Claude AI system, alleging that foreign actors may use it to plan virus-related experiments. The reported activity is described as involving discussions tied to gain-of-function research concepts and dangerous biological materials.
According to the accounts cited by outlets, the foreign research planning reportedly includes reference to multiple pathogens and toxins, including bird flu and chikungunya. Fox News also describes the work as involving researchers outside the United States using Claude to help structure or explore plans related to virus experiments.
The coverage focuses on the nature of the warning and what it implies about AI-enabled assistance to harmful bio-research. Both sources frame the issue as an exposed bioweapon-related risk, but the available information emphasizes Anthropic’s alarm rather than confirming any specific operational attack or completed experiment.
