Policy Design Manager, Conventional Weapons

About the role

Anthropic’s Safeguards team builds the policies, evaluations, and enforcement systems that define how its AI model Claude can be used. As the Policy Design Manager for Conventional Weapons, you will define the boundary between permissible research and prohibited weapons development, and translate that boundary into actionable policy and technical controls.

Key responsibilities

• Own and maintain Anthropic’s conventional weapons policy, clearly defining what model‑supported activities are prohibited.
• Develop threat models and evaluation frameworks that measure how the model could contribute to weapons design, including software and autonomy components.
• Partner with engineering to turn policy into model guardrails, detection systems, and enforcement tooling.
• Serve as the subject‑matter expert for escalations involving weapons‑related content and respond rapidly to emerging risks.
• Communicate policy rationale across product, engineering, legal, and leadership audiences.
• Engage external experts, government and industry partners, and incorporate their input into stronger policy and enforcement.

Required profile

• Deep, applied expertise in weapons systems and ability to translate technical evidence into policy judgments.
• Experience in a defense‑related research lab, agency, or a company that designs weapons systems.
• Proven ability to write clear, operational policy and explain complex technical topics to non‑specialists.
• Understanding of legal frameworks governing weapons and export controls across jurisdictions.
• Comfort with ambiguity and motivation to prevent misuse while supporting legitimate research.

Required skills

• Machine learning and large‑language‑model fundamentals.
• Systems engineering, robotics, autonomy, guidance/navigation/control, sensors, signal processing, aerospace or mechanical engineering, materials, energetics, embedded software.
• Arms export controls (ITAR/EAR) and international arms‑control regimes.
• Design and evaluation of classifiers, including LLM‑based classifiers for high‑consequence categories.
• Trust & safety or product policy experience on a technology platform.

What we offer

• Annual salary range $245,000 – $285,000 USD.
• Competitive benefits, optional equity‑donation matching, generous vacation and parental leave, flexible working hours, and a collaborative office space in San Francisco.