Anthropic announced that it has disrupted attempts to use its Claude models for malicious activities, including research into adapting bird flu to a human-transmittable strain with pandemic potential. The company also detected Chinese distillation attacks, according to a report released on Thursday. The disclosure comes amid increasing concerns about AI misuse and rogue actors. Anthropic said it blocked these efforts as part of its safety measures.
Related Articles
Don't miss out on breaking stories and in-depth articles.