Anthropic warns about the biological risks of artificial intelligence!
Anthropic has revealed cases where its AI models were used to support the development of biological weapons.
Anthropic has revealed that cases were identified in which its artificial intelligence models were used to support the development of biological weapons. In a report published by the company, biological misuse was highlighted as one of the major risks associated with AI.
The report includes examples of users attempting to use the Claude model for dangerous biological activities. In one case, a researcher reportedly tried to use Claude to plan experiments aimed at adapting avian influenza to mammals. In another case, a government-funded researcher was reported to have used AI to create toxins on a computer.
After identifying these cases, Anthropic blocked the relevant accounts and took additional measures to prevent similar activities in the future. The company said that sharing these examples could help improve understanding of biological risks associated with AI and support discussions on how to prevent such misuse.