Head of Policy Design, Societal Harms (AI Safety)

Location: San Francisco, California, United States. Staff are expected to work from one of the offices at least 25% of the time.

Annual salary: $330,000–$395,000 USD.

Company

Anthropic develops reliable, interpretable, and steerable AI systems designed to be safe and beneficial for users and society.

What you will do

• Lead, develop, and grow managers and teams responsible for consumer harm areas including child safety, user well-being, harmful manipulation, and election integrity.
• Coordinate policy decisions across the portfolio and create mechanisms for tracking ownership, rationale, consistency, and implementation.
• Set the strategy for combining model-trained values with policies, detection and enforcement systems, and product interventions.
• Prioritize competing harm areas, make resource tradeoffs, and communicate decisions to leadership.
• Act as the escalation point for high-severity and ambiguous consumer harm decisions, including emerging risks.
• Partner with engineering, data science, product, legal, research, external experts, civil society organizations, and regulators across the model development cycle.

Requirements

• Experience leading teams, including managers or senior specialists, in AI safety, product policy, or a related field.
• Applied expertise in consumer harm areas such as child safety, mental health and well-being, manipulation, or election integrity.
• Strong cross-functional collaboration skills and experience reaching shared decisions across teams.
• Working understanding of frontier model training, fine-tuning, evaluations, launches, consumer products, APIs, and agentic tools.
• Experience translating policy positions into enforceable and measurable mechanisms and explaining decisions to technical, non-technical, and executive audiences.
• Bachelor’s degree or equivalent education, training, or experience in a relevant field; sound judgment in ambiguous, high-consequence situations.

Nice to have

• Subject-matter depth in one or more consumer harm areas from academia, clinical practice, civil society, government, or trust and safety.
• Experience with model training or research teams and generative AI safety systems, including LLM-based classification, evaluation, or enforcement pipelines.
• Experience engaging child safety organizations, election authorities, mental health experts, or regulators.
• Experience using agentic AI tools to scale analysis and operations.

Culture & Benefits

• Collaborative work across research, engineering, policy, and business teams.
• Flexible working hours and an office environment designed for collaboration.
• Competitive compensation, optional equity donation matching, generous vacation, and parental leave.
• Visa sponsorship is available, with immigration lawyer support, although sponsorship cannot be guaranteed for every role or candidate.