Frontier AI Safety Evaluator

Obsidian seeks experienced AI Safety Practitioners to evaluate frontier models on safety, quality, and alignment across grey-area topics. You will assess AI responses, apply safety policies, and contribute to improvements through structured evaluations and feedback.

Responsibilities include reviewing high-risk content, refining RLHF/SFT rubrics, and collaborating with researchers to advance safety benchmarking. A background in journalism, public policy, or related fields is ideal.