Head of Policy Design, Societal Harms

Job summary

Responsibilities

Anthropic's Safeguards organization builds the policies, evaluations, and detection and enforcement systems that define and hold the limits on how Claude can be used. In this role, you'll lead our policy design team, managing the teams responsible for radicalization, child safety, user well-being…

• Important context for this role: some of the work involves exposure to explicit content, including material of a sexual, violent, or psychologically disturbing nature.
• Lead, develop, and grow the managers and teams responsible for the consumer harms portfolio, including child safety, user well-being, harmful manipulation, and election integrity
• Coordinate policy decisions across the portfolio, and build the mechanisms that keep them tracked, consistent, and legible — so stakeholders know what was decided, why, and who…
• Set the strategy for how mitigations built on top of the model — policies, detection and enforcement systems, and product interventions — complement what is trained into the…
• Prioritize across harm areas competing for the same resources, and make those tradeoffs and their rationale clear to leadership

What they're looking for

• Experience leading teams — including managing managers or senior specialists — in AI safety, product policy, or a related field
• Deep, applied familiarity with consumer harm areas such as child safety, mental health and well-being, manipulation, or election integrity, and good judgment about how these…
• A track record of exceptional cross-team collaboration: building durable working relationships with teams you don't control, and getting to shared decisions where ownership is…
• Working understanding of how frontier models are developed and deployed — the training and fine-tuning cycle, evaluations, and launch processes — and how different model…
• Experience translating policy positions into mechanisms that can be enforced and measured, and communicating the reasoning to technical and non-technical audiences, including…