Adolescent Mental Health Clinical Expert (Contract)

About the Role

We’re looking for a Clinical AI Safety Contractor to help evaluate how AI systems respond in mental health and other clinically sensitive situations.

You’ll bring clinical judgment to reviewing AI conversations, identifying safety risks, and helping define what appropriate model behavior looks like. This is a flexible, part-time role for clinicians interested in applying their expertise to AI safety and evaluation.

What You'll Do

• Review and rate AI conversations for clinical safety, appropriateness, and quality

• Identify clinically meaningful risks and model failure modes

• Help create realistic scenarios, evaluation criteria, and scoring rubrics

• Provide expert feedback on how AI should respond in sensitive situations

• Participate in calibration and review sessions with researchers and engineers

• Document important edge cases and emerging risks

Requirements

• Clinical training and professional credentials are required, such as LCSW/LICSW, LMFT, LPC/LPCC/LCPC, clinical psychologist, psychiatrist, MD/DO, or comparable clinical mental health credentials

• Experience working directly with patients or clients in a mental health setting

• Strong grounding in clinical assessment, psychopathology, risk evaluation, or crisis response

• Strong written communication and clinical judgment

• Comfort reviewing sensitive mental health content and maintaining strict confidentiality

No PhD or prior AI experience is required.

Nice to Have

• Experience with crisis intervention, suicide or self-harm assessment, or safety planning

• Experience in adolescent mental health, psychosis, eating disorders, trauma, or other clinically complex areas

• Experience developing clinical rating systems, assessment criteria, or coding frameworks

• Familiarity with AI, digital mental health, trust & safety, or conversational systems

About us

Vals AI builds rigorous evaluations and benchmarks for frontier AI systems. Our work started from NLP evaluation research at Stanford, and today we work across technical and domain-specific areas, including healthcare and mental health. We’ve raised a $5M seed and our team has backgrounds at Stanford, NVIDIA, Meta, Microsoft, Palantir, HRT, Jane Street, and Snorkel.

Further Reading:

• AI tools mostly fumble basic financial tasks, study finds

• The Winners (and Losers) of This New Vibe-Coding Benchmark Will Surprise You

• OpenAI’s Less-Flashy Rival Might Have a Better Business Model

• Meta Announces New AI Model in Major Test of Company’s Ambitions

• DeepSeek’s Sequel Set to Extend China’s Reach in Open-Source A.I.