Head of Policy Design, Societal Harms
San Francisco, CA
Posted 16d ago
Remote Work Policy
On-site
Categories
Applied AI Engineer
About the job
Anthropic's Safeguards organization is responsible for developing policies, evaluations, and enforcement systems to define and manage the limits of Claude's usage. In this leadership role, you will manage the policy design team, overseeing areas such as radicalization, child safety, user well-being, harmful manipulation, and election integrity. Your team will focus on understanding and defining the risks associated with Claude, how they manifest in the real world, and the necessary mitigations. You will collaborate with research, product, and engineering teams to establish boundaries and implement interventions, ensuring safety measures evolve with model capabilities and user behaviors. This role requires a blend of expertise in harm areas and an understanding of frontier AI model development and deployment, with a strong emphasis on cross-functional collaboration and team development.
Responsibilities
- Lead and grow teams responsible for the consumer harms portfolio, including child safety, user well-being, harmful manipulation, and election integrity.
- Coordinate policy decisions across the portfolio, establishing mechanisms for tracking, consistency, and clarity.
- Set strategy for model-based mitigations (policies, detection, product interventions) in conjunction with model training.
- Prioritize across competing harm areas and communicate tradeoffs and rationale to leadership.
- Serve as an escalation point for high-severity and ambiguous consumer harms decisions, including rapid response to emerging risks.
- Partner with engineering, data science, product, legal, and research throughout the model development cycle to ensure consumer harms considerations are integrated.
- Engage external experts, civil society organizations, and regulators, translating insights into policy and enforcement improvements.
Requirements
- 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 (e.g., child safety, mental health, manipulation, election integrity) and their mitigation strategies.
- Proven track record of exceptional cross-team collaboration and building relationships with teams outside of direct control.
- Working understanding of frontier model development and deployment processes (training, fine-tuning, evaluations, launch) and how different environments affect risk and mitigation.
- Experience translating policy positions into enforceable and measurable mechanisms, communicating reasoning to technical and non-technical audiences.
- Sound judgment in ambiguous, high-consequence decisions, with the ability to make calls and escalate appropriately with incomplete information.