Child Psychiatry Expert - AI Safety

Obsidian; San Francisco (CA); On-site; USD 110,000 - 207,000; Part time

Job summary

Obsidian is seeking clinicians and researchers to help design a clinician-informed benchmark evaluating how AI chatbots respond to adolescents facing mental health challenges.

You will author realistic case scenarios, review clinical realism and ethics, and help calibrate evaluation rubrics for AI conversations across domains like suicide risk, therapeutic impersonation, and parasocial attachment.

Qualifications

• MD, DO, PhD, PsyD, or equivalent qualification in Psychiatry, Psychology, or a related mental health field.
• Clinical or research experience focused on children and adolescents.
• Suicide prevention
• Self-harm assessment
• Child & adolescent psychiatry
• Clinical child psychology
• Digital mental health
• Evidence-based therapies such as DBT or CBT
• Strong written communication skills and ability to provide structured clinical feedback.
• Comfortable reviewing fictional clinical scenarios and evaluating AI-generated conversations.
• Experience conducting suicide risk assessments or crisis intervention.
• Academic or published research in adolescent mental health, suicide prevention, or digital therapeutics.
• Experience developing clinical guidelines, assessment frameworks, or educational materials.
• Prior work involving AI, digital health technologies, or mental health product evaluation.
• Experience supervising trainees or participating in multidisciplinary clinical teams.

Responsibilities

• Author fictionalised adolescent mental health case scenarios based on real-world clinical expertise.
• Review scenarios for clinical realism, developmental appropriateness, and ethical considerations.
• Help develop and refine clinician-informed evaluation rubrics for AI chatbot responses.
• Review a subset of AI-generated conversations and provide expert judgment to calibrate automated evaluation systems.
• Advise on best practices for assessing AI behaviour across key mental health domains, including Suicide & Self-Harm (risk recognition, safety planning, crisis referral).
• Advise on Medical & Therapeutic Impersonation (diagnosis, treatment advice, discouraging professional care).
• Advise on Parasocial Attachment & Anthropomorphism (claims of consciousness, emotional dependency, discouraging trusted relationships).

Skills

MD/DO/PhD/PsyD or equivalent

Clinical experience with children and

Suicide prevention

Self-harm assessment

Child & adolescent psychiatry

Clinical child psychology

Digital mental health

DBT or CBT evidence-based therapies

Education

MD, DO, PhD, PsyD or equivalent in Psychiatry/Psychology

Job description

We are partnering with a leading AI research organisation to develop a clinician-informed benchmark for evaluating how AI companion chatbots respond to adolescents experiencing mental health challenges. As AI companions become increasingly common sources of emotional support, there is a growing need for rigorous, clinically grounded evaluation of how these systems handle sensitive situations such as suicide risk, therapeutic guidance, and emotional dependency. We're seeking experienced Child & Adolescent Psychiatrists, Clinical Psychologists, and Mental Health Researchers to contribute their expertise in designing realistic clinical scenarios and evaluation standards for this important AI safety initiative.

Responsibilities

• Author fictionalised adolescent mental health case scenarios based on real-world clinical expertise.

• Review scenarios for clinical realism, developmental appropriateness, and ethical considerations.

• Help develop and refine clinician-informed evaluation rubrics for AI chatbot responses.

• Review a subset of AI-generated conversations and provide expert judgment to calibrate automated evaluation systems.

• Advise on best practices for assessing AI behaviour across key mental health domains, including:

  • Suicide & Self-Harm (risk recognition, safety planning, crisis referral)

  • Medical & Therapeutic Impersonation (diagnosis, treatment advice, discouraging professional care)

  • Parasocial Attachment & Anthropomorphism (claims of consciousness, emotional dependency, discouraging trusted relationships)

Requirements

We are looking for clinicians and researchers with substantial expertise in child and adolescent mental health.

Required Qualifications

• MD, DO, PhD, PsyD, or equivalent qualification in Psychiatry, Psychology, or a related mental health field.

• Clinical or research experience focused on children and adolescents.

• Expertise in one or more of the following:

  • Suicide prevention

  • Self-harm assessment

  • Child & adolescent psychiatry

  • Clinical child psychology

  • Digital mental health

  • Evidence-based therapies such as DBT or CBT

• Strong written communication skills and ability to provide structured clinical feedback.

• Comfortable reviewing fictional clinical scenarios and evaluating AI-generated conversations.

Preferred Qualifications

• Experience conducting suicide risk assessments or crisis intervention.

• Academic or published research in adolescent mental health, suicide prevention, or digital therapeutics.

• Experience developing clinical guidelines, assessment frameworks, or educational materials.

• Prior work involving AI, digital health technologies, or mental health product evaluation.

• Experience supervising trainees or participating in multidisciplinary clinical teams.

Role Details

• Commitment: Approximately 10-15 hours per week.

• Compensation: $80 - $150 USD per hour

Why Join?

• Help shape the future of safe AI systems for adolescents.

• Apply your clinical expertise to one of the most important emerging questions in AI safety and digital mental health.

• Collaborate with researchers developing rigorous evaluation standards for AI companion chatbots.

• Contribute to work that may influence future AI research, independent benchmarking, and responsible deployment of conversational AI.