Nuclear Safeguards AI Red-Teaming Specialist

Mercor; San Francisco (CA); Remote; USD 90,000 - 150,000; Full time

Job summary

Mercor is assembling a panel of nuclear materials and safeguards experts to red-team frontier AI models. The role focuses on writing challenging prompts across benign, dual-use, and adversarial categories, and evaluating model responses against a policy standard.

You will also craft a reference answer with detailed reasoning to justify correctness. This is writing-intensive work requiring clear, written rationales for each judgment and sustained exposure to misuse scenarios in the field.

Qualifications

• Prior technical writing experience or published research is a strong signal.
• Ability to write clear, traceable rationale for non-specialists.
• Experience with red-teaming or security policy concepts is a plus.

Responsibilities

• Write challenging single-turn prompts labeled across benign, dual-use, and adversarial levels.
• Evaluate model responses against a defined policy standard and judge appropriateness.
• Write the reference answer and the technical reasoning explaining why it is correct.

Job description

Mercor is assembling a panel of nuclear materials and safeguards experts to red-team frontier AI models. The role focuses on writing challenging prompts across benign, dual-use, and adversarial categories, and evaluating model responses against a policy standard.

You will also craft a reference answer with detailed reasoning to justify correctness. This is writing-intensive work requiring clear, written rationales for each judgment and sustained exposure to misuse scenarios in the field.