Red-Teaming QA Lead (Remote US Contractor)

About OpenTrain

OpenTrain is the #1 platform for people building careers in AI training and data labeling. We help contributors discover projects, build a unified portfolio of AI training work, and grow their freelance careers with one place to manage proof-of-work.

OpenTrain AI is the hiring and contracting organization for this role. We run and coordinate red-teaming and safety evaluation projects and hire expert contractors to lead work that directly shapes how modern AI behaves.

About AI training and red-teaming

AI training (data labeling and human feedback) is the human side of building AI models — people create, rate, and evaluate examples that teach models to behave safely and usefully. Red-teaming and safety evaluation focus on finding misuse, jailbreaks, and harmful outputs so models can be improved before release.

This work is often remote and flexible, and contributors directly influence model safety by reviewing adversarial prompts, tagging risky outputs, and refining policy-guided rubrics.

The role at a glance

As a Red-Teaming QA Lead you will own quality and consistency across red-teaming and safety-evaluation tasks: auditing submissions, mentoring contributors, maintaining documentation, and escalating recurring or critical issues to keep projects aligned with safety goals.

• Contractor, part-time role focused on QA for adversarial prompts, risk analyses, and safety evaluations.
• Work remotely from the United States and contribute at least 20 hours per week.
• Compensation up to $100/hour.

What you'll do

• Review adversarial prompts, model responses, risk classifications, safety analyses, policy explanations, and vulnerability reports for accuracy, realism, and usefulness.
• Assess trainer and QA submissions against detailed rubrics and provide precise written feedback to improve quality.
• Spot quality issues, flag recurring or critical problems, and escalate where appropriate.
• Communicate project updates, workflow changes, and quality expectations to trainers and QAs.
• Answer contributor questions about risk categories, adversarial strategies, policy boundaries, edge cases, severity, and rubric interpretation.
• Maintain and update style guides, trackers, FAQs, calibration tasks, examples, honeypots, and onboarding materials.
• Support onboarding and training calls for new trainers and QAs to calibrate expectations and quality standards.
• Identify workflow improvements and help scale QA processes for multiple red-teaming projects.

Requirements

• Strong background in AI safety, red-teaming, cybersecurity, trust & safety, content policy, risk analysis, model evaluation, or content moderation.
• Experience evaluating work against detailed rubrics and spotting weak adversarial design, unrealistic scenarios, or unsafe outputs.
• Deep understanding of AI risk categories, jailbreak patterns, harmful-content taxonomies, misuse scenarios, policy interpretation, and safety-evaluation principles.
• Excellent written English communication and the ability to give clear, actionable feedback.
• Comfort working with remote teams and collaboration tools such as Discord, Google Sheets, and Google Docs.
• Experience leading or supporting reviewers, policy analysts, annotators, researchers, or QA teams is strongly preferred.

Schedule, location, and pay

This is a remote contractor role open to candidates located in the United States. The role requires 20+ hours per week and is classified as part-time contract work.

Compensation is up to $100 per hour. Work assignments will focus on text-based evaluation and RLHF-style rating tasks tied to red-teaming and safety projects.

• Location: Remote (United States only).
• Time: 20+ hours per week (part-time contractor).
• Pay: Up to $100/hour.

Who should apply and how it works

You should apply if you are an experienced AI-safety or red-teaming practitioner who enjoys mentoring others, enforcing high-quality rubrics, and improving scalable QA processes. This role is for people who can balance careful judgment with clear, constructive feedback.

To apply, create or update your OpenTrain profile to reflect relevant red-teaming, safety, or QA experience. OpenTrain handles contracting and coordination for this role and will share onboarding details with selected contractors.