Radiological Safety AI Red Team Expert (Remote)
Listing checked September 15, 2026 · pay as published by Mercor
Overview
Neon seeks radiological safety practitioners to red-team frontier AI models on a task-based engagement. You will craft single-turn prompts that sit on the dual-use line, label each one as benign, dual-use, or adversarial, then judge how the model handles the request against a set policy. The work matters because a useful model must answer routine dose, shielding, or isotope questions while refusing requests that seek misuse details. Your domain knowledge decides where that boundary falls, and your writing makes the call clear to people outside radiological safety.
What You'll Do6
- 1Write single-turn prompts in radiological safety that test model boundaries, with each prompt tagged as benign, dual-use, or adversarial.
- 2Review model replies against the program's policy standard and decide whether each response handled the request in the right way.
- 3Draft a reference answer that shows what a sound response looks like and the technical reasoning behind it.
- 4Separate routine professional questions from requests that fish for misuse details, using cases such as isotope therapy, industrial radiography, shielding calculations, or source concealment.
- 5Document the rationale for every judgment in plain language that a non-specialist can follow.
- 6Flag prompts or model replies where the policy line stays unclear so the panel can resolve the ambiguity.
Requirements7
- 1Direct experience with radioactive material responsibility, such as source custody, broad-scope licensing, inventory, transfer, disposal, or security and response duties.
- 2A background that matches one or more of these roles: Radiation Safety Officer for a broad-scope licensee, Health Physicist at a DOE site or national lab, source security or vulnerability assessment specialist, radiological emergency response specialist, or nuclear medicine physicist or cyclotron RSO.
- 3Prior red-teaming work strengthens an application.
- 4Ability to tell a routine professional question from one that seeks misuse guidance, even when both use the same isotope or shielding math.
- 5Strong technical writing. Candidates with published research, prior technical writing, or expert witness work should include a sample or link.
- 6Comfort reading and writing about misuse scenarios in radiological safety over long sessions, with the option to pause or step away.
- 7No reliance on classified, export-controlled, NDA-covered, or prepublication-review material. If you hold such obligations, disclose them in your application so the team can scope the work.
Who Should Apply
The right candidate has held responsibility for radioactive material, not just worked near it, and can explain why a routine dose or shielding question differs from a request that aids misuse. Radiation Safety Officers, health physicists, source security assessors, radiological emergency responders, and nuclear medicine physicists fit this profile. The role is less suitable for generalists without source custody, licensing, or response experience, and for people who cannot produce a written rationale that a non-specialist can follow. Candidates often score low when their prompts do not sit on the dual-use boundary or when their reference answers skip the technical reasoning. A short writing sample makes a difference, so include one if you have it.
Salary Insight
The listed pay is $75 as a one-time payment for this task-based engagement. That amount matches a short, discrete red-teaming contribution rather than ongoing part-time or full-time work. If you expect a recurring rate by the hour or a long project, this role does not offer that.
Location
Compensation
$75 fixed
Required Skills
Application Tip
Name the exact radioactive material responsibility you held, such as broad-scope RSO licence, Category 1 or 2 source security work, or FRMAC response, and attach a short writing sample that walks a non-specialist through a benign versus adversarial call. State any NDA, export-control, or prepublication obligations up front so the team can scope your task.
See NearSkill jobs more often in your search
Application & verification flow
1Instant rubric match
Your resume is scanned against this role’s requirements to check qualification fit.
2Screened before the employer sees it
Only profiles that clear screening are passed on.
3Outcome by email
We notify you at the address on your resume once the screening is reviewed.
Similar open positions
Explore active roles that match your skills and interests.
Mercor
VerifiedRadiation Safety Officer - Remote AI Red Team
A panel of radiological safety experts will red-team frontier AI models. The core task is to test whether a model can spot misuse potential in technical requests while still answering legitimate questions. You write single-turn prompts across benign, dual-use, and adversarial levels, score model replies against a policy standard, and draft reference answers with technical reasoning. The domain sits on a difficult line: the same isotope can support cancer therapy or industrial radiography, and the same shielding calculation can protect a worker or hide a source. The work needs practitioners who have held the licence, because generalists often miss that boundary.
Mercor
VerifiedRadiological Emergency Response Specialist - Remote
This project brings together radiological safety experts to red-team frontier AI models. The panel tests whether a model can judge the misuse potential of a technical request, answering legitimate questions while refusing dangerous ones. Your domain sits on the dual-use line: the same isotope supports cancer therapy and industrial radiography, and the same shielding math protects a worker or hides a source. A model that refuses a routine dose question fails just as much as one that answers a harmful query. Radiological safety practitioners, not generalists, draw that line.
Mercor
VerifiedRadiological Source Security Specialist - Remote
Mercor needs radiological safety practitioners to red-team frontier AI models. You will craft single-turn prompts that test whether a model can separate legitimate technical questions from dangerous misuse, with each prompt tagged benign, dual-use, or adversarial. The work also calls for reviewing model responses against a policy standard and writing reference answers that show the correct call, plus the reasoning behind it. The domain sits on the dual-use line: the same isotope can support cancer therapy or industrial radiography, and the same shielding math can protect a worker or hide a source.
Mercor
VerifiedHealth Physicist - Remote AI Red-Teaming Contract
A project assembles a panel of radiological safety experts to red-team frontier AI models. The core question: can a model judge the misuse potential of a technical request, answer legitimate questions, and refuse requests that pose real harm; a refusal of a routine dose question counts as a failure, just as an answer that should have been refused does. Experts write challenging single-turn prompts in health physics, label each as benign, dual-use, or adversarial, then assess model replies against a defined policy standard. The work also requires reference answers that show the correct response and the technical reasoning behind it. The domain sits on the dual-use line: the same isotope supports cancer therapy and industrial radiography, and the same shielding calculation protects a worker or conceals a source.
Mercor
VerifiedNuclear Safeguards Red Team Expert (Remote)
Neon seeks nuclear domain specialists to pressure-test frontier AI models on misuse judgment. You will craft single-turn prompts across three bands: benign, dual-use, and adversarial. Then you evaluate model replies against a policy standard and write the reference answer with your reasoning. The panel values fuel-cycle depth, safeguards, and nonproliferation work. The task-based remote engagement pays a one-time $75 fee.
Mercor
VerifiedNuclear Security Engineer (Remote, Task-Based)
A panel of nuclear materials and safeguards specialists will red-team frontier AI models for Mercor, testing how well a model judges misuse potential. You write single-turn prompts at three levels: benign, dual-use, and adversarial, then judge model replies against a policy standard and draft the reference answer with technical reasoning. The work sits on the dual-use line where a material-balance calculation can close an inspector's account or hide a gap. Expect sustained reading and writing about misuse scenarios, with briefings and freedom to pause without penalty.