Nuclear Safeguards Red-Teaming Specialist

Mercor is assembling a panel of nuclear domain experts to red-team frontier AI models, testing whether a model can correctly judge the misuse potential of a technical request. You will write challenging prompts labeled across benign, dual-use and adversarial levels, evaluate responses against policy standards, and draft reference answers with the underlying rationale.

This role emphasizes clear technical writing and exposure to safeguards concepts, with the ability to explain decisions to