Nuclear Safeguards & AI Red-Teaming Expert

Mercor; Greater London; On-site; GBP 55,000 - 75,000; Full time

Job summary

Mercor is assembling a panel of nuclear domain experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones.

This role is writing-intensive: you will write prompts across benign, dual-use, and adversarial levels; evaluate the responses against policy standards; and draft the reference answer with technical reasoning for why a response is

Qualifications

• Red-teaming prompts writing (benign, dual-use, adversarial)
• Judging model responses against policy standards
• Drafting reference answers with technical rationale

Responsibilities

• Write challenging single-turn prompts across three levels
• Evaluate model responses and judge handling
• Write reference answers with rationale

Skills

Red-teaming experience

Nuclear domain knowledge

Technical writing

Education

Safeguards Analyst background

Nuclear Engineer background

Nuclear Security knowledge

Job description

Mercor is assembling a panel of nuclear domain experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones.

This role is writing-intensive: you will write prompts across benign, dual-use, and adversarial levels; evaluate the responses against policy standards; and draft the reference answer with technical reasoning for why a response is