Nuclear Safeguards Red-Team Analyst

Obsidian seeks contributors to assemble a panel of nuclear domain experts to red-team frontier AI models. You will craft prompts across benign, dual-use, and adversarial levels and assess model responses against policy standards. A key deliverable is a reference answer with thorough technical reasoning.

The role emphasizes safeguards expertise, precise writing, and the ability to explain judgments to non-specialists. Prior research or publication in this field is highly valued.