Claude helped with research that could aid bioweapons, according to @AnthropicAI. Some work passed its filters because the same science can support medicine. The report establishes neither harmful intent nor an imminent threat. https://digg.com/tech/v5k9vkcm?utm_sourc…