AI Safety & Red Team Expert at Mercor

AI Safety & Red Team Expert | $29–45/hr | Worldwide Remote

Mercor is assembling a team of adversarial AI experts to probe, stress-test, and improve the safety of cutting-edge AI models. This role is ideal for professionals with backgrounds in AI red teaming, cybersecurity, or socio-technical risk analysis. Native fluency in both English and Portuguese (global, excluding Brazilian Portuguese) is required.

Why This Role Exists

The safest AI is one that has already been challenged. Mercor's red team generates the adversarial human data that helps customers build more robust, trustworthy AI systems. You'll surface vulnerabilities that automated tests miss and deliver reproducible findings that drive real safety improvements.

Note: This project involves reviewing AI outputs on sensitive topics such as bias, misinformation, and harmful behaviors. All work is text-based. Participation in higher-sensitivity projects is optional and supported by clear guidelines and wellness resources. Topics are communicated in advance.

What You'll Do

• Red team conversational AI models and agents — including jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Generate high-quality human data: annotate failures, classify vulnerabilities, and flag systemic risks
• Apply structured frameworks, taxonomies, and benchmarks to ensure consistent testing
• Produce reproducible reports, datasets, and attack case documentation for customer use

Who You Are

• Experienced in AI adversarial work, cybersecurity, or socio-technical probing
• Naturally curious and adversarial — you instinctively push systems to their limits
• Structured in your approach, using established frameworks rather than ad hoc methods
• A clear communicator who can explain risks to both technical and non-technical audiences
• Adaptable and comfortable moving across diverse projects and customer environments

Nice-to-Have Specialties

• Adversarial ML: Jailbreak datasets, prompt injection, RLHF/DPO attacks, model extraction
• Cybersecurity: Penetration testing, exploit development, reverse engineering
• Socio-technical risk: Harassment/disinfo probing, abuse analysis, conversational AI testing
• Creative probing: Psychology, acting, or writing skills applied to unconventional adversarial thinking

What Success Looks Like

• You uncover vulnerabilities that automated testing overlooks
• You deliver reproducible artifacts that directly strengthen customer AI systems
• Evaluation coverage expands — more scenarios tested, fewer surprises in production
• Mercor customers trust the safety of their AI because you've already probed it like an adversary

Why Join Mercor

• Build hands-on experience in human data-driven AI red teaming at the frontier of safety
• Play a direct role in making AI systems more robust, safe, and trustworthy