Anthropic Hiring Policy Design Manager to Set AI Use Limits in Weapons

Two days after CEO Dario Amodei's essay on AI risks, Anthropic posted a policy design manager role in its Safeguards team. The hire will define acceptable use boundaries for Claude in conventional weapons, including autonomous targeting systems.

• Job posted September 14, 2026; Amodei's essay on September 12
• Role covers all weapons classes, including autonomous systems
• Anthropic previously refused to give Pentagon models for autonomous weapons
• Reuters reported Anthropic tools were used in US attack on Iran