Behavioral Health Expert, AI Safety and Model Evaluation

$45–70/hr

rate as listed

Workload: 20+ hrs/week
Engagement: Project-based
Location: US
Field: Sciences

Evaluate AI responses to sensitive conversations about relationships, emotional wellbeing, personal beliefs and difficult decisions. You will identify unsafe or unbalanced patterns, write evidence-based rationales and help develop evaluation rubrics and test scenarios.

The work

You will review conversations between users and AI models, then assess whether the model remains neutral, supportive and appropriately bounded. Day-to-day tasks include flagging sycophancy, taking sides, moralizing, reinforcement of distorted or unfounded beliefs, and responses that drift into clinical or directive advice. You may also draft reference responses, develop evaluation rubrics and design non-crisis test conversations involving relationships, family dynamics, emotional wellbeing, spirituality and unconventional beliefs.

What the screening emphasizes

The platform is likely to verify at least three years of relevant professional experience and probe how you apply behavioral health principles to ambiguous examples. Strong candidates can distinguish validation from agreement, explain when a response oversteps its role, and ground judgments in concepts such as cognitive distortions, healthy boundaries, client-centered practice and motivational interviewing. Clinical licensure is valued but not required, while experience in AI safety, applied ethics, trust and safety, content policy, relationship counseling, spiritual care, misinformation research or AI evaluation is helpful.

Schedule and collaboration

The minimum commitment is 20 hours per week on weekdays, with the option to increase to as many as 40 hours per week. The work includes collaboration with researchers and other experts to calibrate standards and document decisions consistently. The listing does not specify whether the position is remote or whether most tasks are completed asynchronously, so candidates should confirm location, meeting and scheduling expectations during the process.

Employment

This is a W-2 position with Cincinnatus LLC, which serves as the employer of record and may place the successful candidate with a leading AI lab as part of its extended workforce.

Pay band: $45–70/hr

What they look for

At least 3 years of professional mental health, counseling or social services experience
Degree in a relevant field or equivalent professional mental health experience
Behavioral health judgment across sensitive, non-crisis conversations
Knowledge of sycophancy, cognitive distortions, healthy boundaries and client-centered approaches
Precise written evaluation and rubric development
Neutral, nonjudgmental reasoning across diverse beliefs and relationships

Included with this listing

Interview prep — 10 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: This verifies the required experience and whether the candidate has directly supported people in mental health, counseling or social services settings rather than only studying the field.
A strong answer: Give a concise chronology with role titles, settings, dates or durations, populations served and the kinds of support or judgment you personally provided. Clearly separate direct practice from research, administration or volunteer work.
Where people lose it: Candidates lose points by naming broad fields such as counseling or advocacy without establishing three years of professional work or clarifying their direct responsibilities.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.