Red-Teaming Quality Assurance Lead

About OpenTrain

OpenTrain is the #1 platform for people building careers in AI training and data labeling. We help freelancers discover projects, build a single portfolio of AI training experience, and grow a durable freelance career—all from one place.

We directly hire and contract contributors for AI training roles. Working with OpenTrain gives you long-term visibility for your work and access to projects that shape how modern AI systems behave.

Why AI Training and Red-Teaming Work Matters

AI training (data labeling, annotation, and human feedback) is the human side of building AI: people create, review, and rate examples that teach models to respond safely and usefully. Red-teaming and safety evaluation are critical to finding vulnerabilities, misuse paths, and policy gaps before models are deployed.

These roles are often remote and flexible, and they let you directly influence model behavior while developing a visible portfolio of safety and evaluation work.

The Role

We are hiring a Red-Teaming Quality Assurance Lead to oversee quality, consistency, and contributor performance across red-teaming and safety evaluation projects. You will evaluate AI-generated safety work and trainer/QA submissions against detailed rubrics, then provide clear, actionable written feedback.

This is an hourly contract role, remote (United States hiring), requiring 20+ hours/week and paying $100 per hour.

What You'll Do

• Review adversarial prompts, model responses, risk classifications, safety analyses, policy explanations, and vulnerability reports for accuracy, realism, and usefulness.
• Spot-check trainer and QA submissions, escalate recurring issues, and reinforce quality standards across projects.
• Provide precise written feedback against project guidelines and rubrics—focus on rubric adherence, instruction-following, formatting, and scenario realism.
• Communicate updates, workflow changes, and review expectations to trainers and QAs in remote collaboration channels.
• Support onboarding, documentation, FAQs, trackers, calibration tasks, and process improvements.
• Help manage contributor activation and follow-up when team members are inactive or inconsistent.

Requirements

You must have strong written English communication, excellent attention to detail, and experience reviewing red-teaming or safety evaluation work. This role expects comfort evaluating work against detailed rubrics and providing clear corrective feedback.

• Proven experience reviewing AI red-teaming, safety evaluations, or rubric-based QA work.
• Ability to assess adversarial prompts, risk analyses, and trainer/QA submissions for quality and policy alignment.
• Familiarity with adversarial prompting, jailbreak patterns, misuse scenarios, and harmful-content taxonomies.
• Experience with remote collaboration tools such as Discord, Google Sheets, Google Docs, trackers, dashboards, or project management systems.
• Available for 20+ hours per week and eligible to work as a US-based contractor.

Helpful Background

The following are not strictly required but will help you succeed in this role and advance within OpenTrain.

• Background in Computer Science, Cybersecurity, AI Safety, Trust & Safety, Public Policy, Psychology, Linguistics, Law, Security Studies, or Risk Analysis.
• Experience with LLM evaluation, content moderation QA, policy QA, or other rubric-based review workflows.
• Familiarity with prompt injection, social engineering, fraud, self-harm safety, extremist content, misinformation, privacy risk, illicit behavior, bias, and model refusal behavior.

Pay, Logistics, and How It Works

This is a contract, part-time position for US-based contractors. Compensation is $100 per hour. Typical commitment is 20+ hours/week; scheduling is flexible within project needs.

You will review text-based evaluation tasks (label type: EVALUATION_RATING) and work inside shared collaboration channels and trackers. OpenTrain hires and contracts directly—use your OpenTrain profile to apply and track proof of work.

• Employment type: Contractor, Part-time.
• Work location: Remote (United States hiring only).
• Payment: Hourly, USD $100/hr.
• Data type: Text; Label types: Evaluation rating tasks.