Nuclear Safeguards & AI Risk Red-Teamer

Mercor; Greater London; On-site; GBP 60,000 - 90,000; Full time

Job summary

Mercor is assembling a panel of nuclear materials and safeguards experts to red-team frontier AI models. You will write challenging prompts across three levels and evaluate model responses against a defined policy standard.

You will also craft the reference answers and the technical rationale explaining why the responses meet the policy criteria, with emphasis on distinguishing routine questions from misuse scenarios.

Qualifications

• This is writing-intensive work requiring a clear, written rationale for each judgment.
• Prior technical writing or published research is a strong signal; samples or links are welcome.
• Ability to read and write about misuse scenarios in the field over sustained periods.

Responsibilities

• Write prompts labeled across benign, dual-use, and adversarial levels.
• Evaluate model responses against a policy standard and judge correctness.
• Write the reference answer and technical reasoning for why it is correct.

Skills

Technical writing

Analytical reasoning

Education

Export control licensing / dual-use classification

Treaty and agreement implementation

Proliferation pathway and programme analysis

Sanctions, procurement network analysis

Open-source and technical analysis of nuclear programmes

Job description

Mercor is assembling a panel of nuclear materials and safeguards experts to red-team frontier AI models. You will write challenging prompts across three levels and evaluate model responses against a defined policy standard.

You will also craft the reference answers and the technical rationale explaining why the responses meet the policy criteria, with emphasis on distinguishing routine questions from misuse scenarios.