Product Manager, Safeguards Rare Harms

About the role

Anthropic is seeking a Product Manager to lead its Safeguards team, focusing on protecting users from the risks associated with powerful AI systems. The role involves designing, developing, and deploying safety mechanisms and user‑facing experiences for frontier models across multiple platforms.

Key responsibilities

• Define and embed safety‑by‑design principles for Anthropic’s models, APIs, and cloud‑based products.
• Write and maintain safety evaluation frameworks and communicate findings externally.
• Prioritize initiatives by balancing business impact, technical feasibility, and risk mitigation.
• Collaborate with policy, enforcement, research, and engineering teams to align safeguards across the product lifecycle.
• Monitor the AI ecosystem to anticipate deployment risks and adversarial threats.
• Develop metrics to surface performance gaps and guide future roadmap decisions.

Required profile

• Proven ability to make technical trade‑off decisions in cross‑functional environments.
• Experience working with AI/ML researchers, policy experts, and software engineers to build safety systems.
• Strong understanding of user interactions with AI products and associated safety concerns.
• Comfort with ambiguous, fast‑moving settings and a commitment to AI safety.

Required skills

• Safety evaluation design and implementation.
• Metrics development for AI risk assessment.
• Cross‑functional collaboration across policy, research, and engineering.