LLM-based classification systems Jobs

Anthropic

As a Safeguards Analyst on the User Well-being team, you will support the design and deployment of mental health guardrails. This involves iterating on detection systems, managing review queues, evaluating new interventions, and monitoring existing ones. The work translates expert clinical guidance, data analyses, and constraints into concrete changes for detection, review, and response. The team addresses harms including suicide, self-harm, disordered eating, AI sycophancy, and emotional dependence on AI, with potential to expand into broader user well-being enforcement. Safety is central to the mission, and this role will help shape policy enforcement for harmless, helpful, and honest user interactions with AI products.

$245k - $285k