AI Red-Teamer — Adversarial AI Testing; English & Arabic

Mercor is building an elite red team to make AI safer. They need experts who probe AI models with adversarial inputs, surface vulnerabilities, and generate data that strengthens AI systems. Native-level fluency in English and Arabic required.

What you’ll do

• Red team conversational AI models and agents: jailbreaks, prompt injections, misuse cases, bias exploitation, multi-turn manipulation
• Generate high-quality human data: annotate failures, classify vulnerabilities, flag systemic risks
• Follow structured taxonomies, benchmarks, and playbooks for consistent testing
• Produce reproducible reports, datasets, and attack cases customers can act on

What they’re looking for

• Prior red teaming experience (AI adversarial work, cybersecurity, socio-technical probing)
• Structured approach: frameworks and benchmarks, not random hacks
• Clear communication to technical and non-technical stakeholders
• Native-level fluency in English and Arabic
• Nice to have: Adversarial ML, penetration testing, RLHF/DPO attacks, creative probing

Compensation

• Pay: $32.25/hour
• Type: Hourly contract, full-time or part-time
• Location: Remote
• Flexible schedule, work on your own time
• Weekly payments via Stripe or Wise

How to apply

Apply through Mercor. Complete a screening interview and start contributing to AI safety projects.