Anthropic has revealed that it disrupted several attempts to misuse its Claude AI models for potentially dangerous purposes over the past eight months, including research that could have contributed to biological weapons development. The disclosure, made in a report published on Thursday and covered by the New York Times, underscores growing concern about how powerful AI systems could be exploited for harmful ends, even as the same technology enables legitimate scientific progress. It comes as debate intensifies over AI safety, following warnings from a whistleblower that Anthropic's technology could pose existential risks to humanity.

The report found that scientists had used Anthropic's platforms to conduct research with potential biological weapons applications, though the company said it could not determine whether the work was intended for legitimate or malicious purposes and chose to act cautiously regardless. Separately, it identified suspected Chinese and Iranian state actors using Claude to surveil dissident and diaspora communities, while Russian state media reportedly used the chatbot to generate propaganda, including fabricated claims about an election in Moldova. Anthropic acknowledged the dual-use dilemma inherent in biological research, noting it can lead to breakthroughs such as vaccines but also carries risks of misuse for creating dangerous pathogens.