About this role
In this hourly, remote contractor role, you will work as a Red-Teaming Quality Assurance Lead (QAL) to oversee quality, consistency, and trainer performance across AI red-teaming and safety-evaluation projects. You will review AI-generated safety evaluations, adversarial prompts, risk analyses, and trainer/QA work; evaluate output quality against project guidelines; provide precise written feedback; and ensure that all contributors follow the expected quality standards.
You will assess work for risk identification, adversarial reasoning, policy awareness, safety taxonomy alignment, prompt quality, scenario realism, vulnerability coverage, clarity, formatting, instruction-following, and adherence to project-specific rubrics. You will spot recurring quality issues, communicate updates to trainers and QAs, support onboarding, maintain documentation, and help activate contributors who are not working consistently. This role requires strong AI safety/red-teaming judgment, strong English communication skills, excellent attention to detail, structured communication, and the ability to manage quality workflows across remote expert teams.
This role is with SME Careers, a fast-growing AI Data Services company and subsidiary of SuperAnnotate, delivering training data for many of the world’s largest AI companies and foundation-model labs. Your red-teaming quality leadership will directly help improve the world’s premier AI models by ensuring that safety training data is realistic, nuanced, policy-aligned, well-documented, and useful for identifying model vulnerabilities.
Selection process involves an AI interview, a domain-specific task, and an interview with a recruiter.
Important:
There is no immediate project for this role; however, if qualified, you will be among the first experts we reach out to when relevant opportunities arise. This will also provide you with access to future projects available through our expert network.
Requirements
- Bachelor’s, Master’s, or professional experience in Computer Science, Cybersecurity, AI Safety, Trust & Safety, Public Policy, Psychology, Linguistics, Law, Security Studies, Risk Analysis, or a related field.
- Strong grasp of the English language to follow project guidelines, communicate with teams, and provide clear written feedback.
- 3+ years of experience in AI safety, red-teaming, cybersecurity, trust and safety, content policy, risk analysis, adversarial testing, model evaluation, content moderation, or related workflows.
- Strong understanding of AI risk categories, adversarial prompting, jailbreak patterns, harmful-content taxonomies, misuse scenarios, policy interpretation, model behavior, and safety evaluation principles.
- Ability to evaluate red-teaming content against detailed rubrics and identify issues such as weak adversarial design, unrealistic scenarios, poor risk categorization, policy misinterpretation, unsafe outputs, or superficial vulnerability testing.
- Familiarity with areas such as prompt injection, social engineering, cybersecurity abuse, fraud, self-harm safety, extremist content, misinformation, privacy risk, illicit behavior, bias, and model refusal behavior is preferred.
- Experience leading or supporting remote teams of red-teamers, reviewers, policy analysts, annotators, researchers, or QAs is strongly preferred.
- Comfortable working in fast-moving remote environments using tools such as Discord, Google Sheets, Google Docs, trackers, dashboards, and project management systems.
- Highly detail-oriented and organized, with the ability to maintain style guides, FAQs, trackers, onboarding materials, calibration tasks, and documentation.
- Experience with AI training, LLM evaluation, safety evaluations, content moderation QA, policy QA, or rubric-based review is a strong plus.
Responsibilities
- Quality monitoring: Spot-check red-teaming items, identify quality issues, provide ongoing feedback through DMs, and escalate recurring or critical issues.
- Safety and red-team review: Evaluate adversarial prompts, model responses, risk classifications, safety analyses, policy explanations, and vulnerability reports for accuracy, realism, and usefulness.
- Trainer and QA communication: Update trainers and QAs on Discord about new item guidelines, project changes, workflow updates, quality expectations, and red-teaming-specific review standards.
- Question handling: Respond to trainer/QA questions clearly and promptly, especially around risk categories, adversarial strategy, policy boundaries, edge cases, severity, and rubric interpretation.
- Trainer/QA activation management: DM contributors who are inactive or not working, encourage activation, track follow-ups, and flag availability issues when needed.
- Documentation: Create and maintain red-teaming project documentation, including style guides, trackers, FAQs, quality notes, examples, honeypots, calibration tasks, and onboarding materials.
- Onboarding and training: Schedule and run onboarding/training calls with trainers and QAs to explain project expectations, workflows, rubrics, quality standards, and red-teaming-specific review requirements.
- Quality alignment: Ensure all trainers and QAs apply red-teaming and safety-review guidelines consistently and understand updates as projects evolve.
- Risk review: Flag unsafe, low-quality, unrealistic, policy-inconsistent, or insufficiently documented red-team items.
- Process improvement: Identify recurring quality gaps, propose workflow improvements, and help build scalable QA processes for AI red-teaming projects.
Skills
- LLM Evaluation
- AI red-teaming
- Safety Taxonomies
- Adversarial Testing
- AI Safety
- Trust And Safety
- Risk Assessment
- Prompt Engineering
- Trust & Safety
- Trainer Feedback
- Policy Evaluation
Additional skills
- AI Safety
- Red Teaming
- Adversarial Testing
- LLM Evaluation
- Model Safety Evaluation
- Prompt Injection
- Jailbreak Testing
- Adversarial Prompting
- Trust & Safety
- Content Moderation
- Content Policy
- Policy Interpretation
- Risk Analysis
- Safety Rubrics
- Quality Assurance (QA)
- QA Auditing
- Calibration & Inter-Rater Reliability
- Harmful Content Taxonomies
- Misuse Scenario Analysis
- Cybersecurity
- Social Engineering
- Fraud & Scam Detection
- Privacy Risk Assessment
- Misinformation Analysis
- Extremism & Terrorism Content
- Self-Harm Safety
- Bias & Fairness Evaluation
- Technical Writing
- Remote Team Leadership
- Google Sheets
