Safeguards Enforcement Analyst (User Well-being)

Location: San Francisco, CA; New York City, NY; or Washington, DC. Hybrid work is required, with staff expected to be in an office at least 25% of the time.

Annual salary: $245,000–$285,000 USD

Company

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

What you will do

• Design and execute mental health safeguard interventions, define success metrics, and curate evaluation datasets.
• Partner with Engineering and Data Science to build, tune, and validate detection models, including threshold-setting and precision-recall tradeoffs.
• Monitor intervention and detection-system performance over time and review flagged content to improve enforcement policies.
• Support product features that connect users with crisis resources, collaborating with Product, Legal, and external partners on referral pathways and user-facing content.
• Identify policy gaps from real-world scenarios and incorporate emerging AI policy and mental-health research into workflows.

Requirements

• Experience in trust and safety, product policy, content moderation, or a related field involving mental health, suicide, self-harm, or other well-being harms.
• Experience designing experiments, evaluations, or measurement studies and translating policy definitions into measurable rubrics or classification criteria.
• Experience coordinating content-review operations, quality assurance, and workflow management.
• Proficiency in SQL or other data-analysis tools, plus experience with generative AI products and effective prompts for review, classification, or evaluation.
• Strong judgment in ambiguous, high-consequence cases and ability to communicate emerging risks to cross-functional stakeholders.
• Bachelor’s degree or equivalent education, training, and experience in a relevant field.

Nice to have

• Subject-matter expertise in mental health through academia, clinical practice, crisis intervention, trust and safety, or a related setting.
• Experience building or evaluating LLM-based classification systems.
• Experience using agentic tools such as Claude Code to scale analysis or automate recurring work.
• Experience working in crisis support.

Culture & Benefits

• Collaborative environment focused on large-scale research and trustworthy AI.
• Flexible working hours and an office-based collaboration model.
• Competitive compensation, optional equity donation matching, generous vacation, and parental leave.
• Visa sponsorship may be available, with immigration-lawyer support, depending on the role and candidate.