You'll author and review evaluation tasks that test whether a model can tell a safe request from an unsafe one, and write the expert analysis behind each call: what makes a prompt acceptable, where it crosses a line, and why.

Why apply

• $150/hr equivalent, paid at $85 per approved task.
• Remote and flexible: task-based work on your own schedule.
• Use your expertise to directly shape how AI models handle safety-critical questions in your field.

• Author prompts and scenarios that probe the boundary between safe and unsafe requests in your field.
• Review and classify prompts written by other experts as safe or unsafe.
• Write the domain analysis behind each judgment: why a prompt is acceptable or where it crosses a line.
• Revise your work based on structured reviewer feedback.

Requirements

• Master's degree or higher in materials science, materials engineering, metallurgy, or a closely related discipline.
• Citizenship of a Five Eyes country (United States, United Kingdom, Canada, Australia, or New Zealand).
• Physically located in one of those countries while doing the work.
• Able to explain the technical reasoning behind a safety judgment clearly and precisely.

• Citizen of a Five Eyes country: United States, United Kingdom, Canada, Australia, or New Zealand.
• Physically located in one of those five countries while doing the work.
• Master's degree or higher in materials science, materials engineering, metallurgy, or a closely related discipline.

Preferred qualifications

• Prior experience in AI evaluation, red teaming, or data annotation work.
• PhD, or professional experience in metallurgy, advanced or energetic materials, nuclear materials, or national laboratory research.

Work arrangement: Fully remote

Applicant eligibility: 5 eligible countries

Citizenship of a Five Eyes country (United States, United Kingdom, Canada, Australia, or New Zealand).

$150/hr.

Pay: $150/hr equivalent, paid at $85 per approved task. Remote, flexible, task-based.