Child and Online Safety Expert

About OpenTrain

OpenTrain is the leading platform for building careers in AI training and data labeling. We help professionals discover projects, consolidate proof-of-work, and grow a durable freelance practice contributing domain expertise to next-generation AI systems.

OpenTrain AI is the hiring and contracting organization for this role. Creating an OpenTrain account is free and gives you one place to manage applications, show experience, and find matched work in AI training.

About AI training and why this work matters

AI training (also called data labeling or human feedback work) is the human side of building safe, useful AI: experts create the examples, taxonomies, and evaluation standards models learn from. This role directly shapes how AI recognizes and responds to risks affecting young people.

Contributors typically work remotely, on flexible schedules, and can apply clinical knowledge to cutting-edge safety systems used across products and research.

The role

We’re hiring Child and Online Safety Experts to provide clinical and operational domain expertise for evaluation and RLHF-style tasks focused on youth mental-health safety and digital risk. You will help define what harmful content affecting minors looks like, how to detect it, and how AI systems should escalate or respond.

This is a contractor, part-time role. Work is remote and worldwide; all contributors must be proficient in English and available for roughly 20+ hours per week.

• Employment type: Contractor, Part-time
• Time commitment: 20+ hours/week (flexible scheduling)
• Work type: Remote, worldwide (English required)

What you’ll do

• Develop taxonomies for self-harm, suicide prevention, eating disorders, and emotional dependency in digital contexts
• Establish clinical and digital-safety standards for youth and adolescent support
• Create detection frameworks to identify, flag, and evaluate harmful online content and advice
• Design and refine clinical response rubrics and escalation protocols for at-risk individuals
• Provide clear written and verbal feedback to improve AI safety tools and evaluation frameworks
• Contribute expertise on child safety, online exploitation prevention, and CSAM-adjacent policy for digital environments

Requirements

• Licensed therapist, psychiatrist, clinician, or mental-health professional with direct crisis-care experience
• Demonstrated background in adolescent mental health, suicide prevention, eating-disorder intervention, or closely related work
• Clinical experience with risk assessment and intervention in digital or online settings
• Experience evaluating child online exploitation prevention initiatives or CSAM-adjacent policies
• Strong written and verbal communication skills in English

Helpful background and ideal candidates

We welcome applicants who bring advanced degrees, specialized certifications, or equivalent operational experience in mental health, child safety, crisis response, or related domains. A recognized track record of impact in clinical, crisis-response, or safety-evaluation communities is a plus.

• Advanced degree or specialized certification in mental health or child safety (helpful but not required if operational experience is equivalent)
• Prior work influencing safety policy, clinician-facing protocols, or evaluation metrics for online harm

Compensation, data type, and how to apply

Compensation is paid hourly at $50–$90 USD per hour depending on experience and assignment. This work is paid per hour as a contractor relationship with OpenTrain AI.

You will work on text-centered evaluation and RLHF-style labeling tasks (data type: TEXT; label types: EVALUATION_RATING, RLHF). Assignments typically involve writing and reviewing taxonomies, annotation guidelines, and response rubrics rather than high-volume micro-labeling.

• Pay: USD $50–$90 per hour (hourly contractor pay)
• Data & labels: TEXT data; EVALUATION_RATING and RLHF-style work
• To apply: create or update your OpenTrain profile, confirm credentials/licensing, and submit your application through OpenTrain AI