AI Safety Push: Anthropic Hires Manager for Chemical, Explosive Threat Risks

• Anthropic, a US-based AI firm, hired a Policy Manager to address chemical weapons and high-yield explosives risks in AI systems.
• The role involves shaping how AI handles sensitive information and evaluating AI model capabilities related to chemical weapons and energetic materials.
• Applicants need 5+ years experience in chemical weapons or explosives defense with deep expertise in related areas.
• Other firms like OpenAI have also recruited for similar "Responsible AI" roles to mitigate catastrophic risks from frontier AI models.
• Experts warn this approach could inadvertently provide AI tools with weapon-related information, despite Anthropic's co-founder Dario Amodei's stance against AI use in war. A BBC report indicates Anthropic's Claude is still deployed by the US in the US-Israel Iran war, despite the company's previous legal action against the US government regarding military use.