Nuclear Engineering & Safeguards Experts for Red Team

Neon is seeking nuclear engineering and safeguards experts to participate in AI safety testing through red team exercises. As part of this remote, task-based role, you'll write detailed prompts that test how AI models handle sensitive nuclear information. Your work involves crafting scenarios across multiple difficulty levels—from straightforward technical questions to complex cases where misuse risks are present—to evaluate whether models can distinguish legitimate inquiry from potentially dangerous requests. This requires deep domain knowledge to authentically assess model behavior in your area of expertise.

This position suits advanced professionals with strong nuclear engineering or safeguards backgrounds who can articulate the nuances of safety protocols and proliferation concerns. The role is fully remote and structured around focused writing tasks rather than ongoing employment. Compensation is listed at $65–$75 per hour for approximately 40 hours of work per week. You'll contribute directly to frontier AI development by helping identify gaps in model safety before deployment.