What it means
Deliberately trying to make a model produce harmful, unsafe or wrong output, so the developer can find and fix weaknesses before release. Roles range from general prompt writing to specialists in security, chemistry, biology or law.
Used in: Red-teaming and safety roles.
In open roles right now
91 of the open roles we list mention “red-teaming” in the title or description, on 14 platforms (as of 23 October 2026). Of those that state an hourly rate (28 distinct postings), the median is $65/hr, and the middle half pay $50 to $85.
Most of them on: Alignerr (28), Mercor (24), SuperAnnotate (9), Handshake AI (6), mpathic (6), Invisible (4), and 8 more.
Cyber Threat Intelligence SMEmicro1·Location not stated
$70–110/hr1 day ago
AI/ML Evaluation Reviewer- Healthcare AI (Contract)mpathic·Location not stated
Pay set per project4 days ago
AI Benchmark & Red-Teaming SpecialistTuring·United States·Full-time
Pay set per project7 days ago
AI Red Teamer (Remote)Handshake AI·United States·Full-time
$32–95/hr18 days ago
Behavioral Health Expert, AI Safety and Model EvaluationMercor·United States·Part-time
$45–70/hr21 days ago
Related terms
Adversarial promptA prompt written to make a model fail: to give a wrong, unsafe or rule-breaking answer.JailbreakA prompt that gets a model to ignore its safety rules, for example through role-play, a hypothetical, or a harmful request split into harmless-looking parts. ← Quality scorePrevious term, A to Z. Response rewriting →Next term, A to Z. The whole glossaryEvery term on one page, A to Z.Open →