Anthropic Threat Report Exposes Attempts To Build Biological Weapons With Claude

Anthropic has disclosed that several scientists and state-backed research groups attempted to use Claude AI to assist with potentially dangerous biological research. According to a detailed threat report, the company identified and blocked these misuse attempts, which could have supported the creation of biological weapons. The report outlines five specific case studies occurring since late last year, confirming that Anthropic successfully intervened to prevent the AI from facilitating these presumed scientific or nefarious biological threats.