Radiological AI Safety Evaluators

AI Safety Evaluation Project | Remote | Contract Locations: US, Canada, UK, Europe, and Australia We are seeking qualified scientific and technical professionals to support an AI safety evaluation project focused on radiological, nuclear, and CBRN-related content. The work involves reviewing AI-generated prompts and responses, evaluating safety risks, identifying dual-use content, and assessing technical accuracy and responsible handling. Tier 1
- First-Pass Annotators Bachelor's or Master's degree in a relevant STEM or technical field. Strong backgrounds may include Physics, Nuclear Engineering, Nuclear Physics, Radiological Science, Radiation Physics, Health Physics, Chemistry, Chemical Engineering, Materials Science, or related engineering disciplines. Relevant professional experience in radiation, nuclear technology, hazardous materials, laboratory science, engineering, or safety may also qualify. Tier 2
- Senior SMEs and Auditors PhD or equivalent advanced industry or professional expertise in a relevant field. Relevant backgrounds include Nuclear Engineering, Nuclear Physics, Physics, Radiological Science, Radiation Physics, Health Physics, Chemistry, Chemical Engineering, CBRN-related science and engineering, or closely related fields. Strong candidates may have professional experience in nuclear or radiological operations, radiation safety, nuclear security or safeguards, CBRN preparedness, hazardous-materials safety, scientific research, national laboratories, defense, or regulatory environments. Who we're looking for We value real-world subject-matter expertise, analytical judgment, and strong attention to detail. Experience in AI evaluation, AI safety, scientific research, technical review, or quality assurance is a plus. CBRN experience is welcome, including expertise in Chemical, Biological, Radiological, or Nuclear domains. Remote and contract, within the eligible locations listed above. Responsibilities
- Tier 1: review AI prompts and model responses
- Tier 1: classify content as Benign, Dual-Use, or Malicious using project guidelines
- Tier 1: evaluate and rank responses for safety, and give a clear rationale for each
- Tier 1: identify the legitimate and the potentially harmful components of dual-use scenarios
- Tier 2: audit and QA Tier 1 annotations, and review complex or flagged cases
- Tier 2: validate classifications, rankings, rationales, and reference responses
- Tier 2: ensure evaluations meet appropriate scientific, technical, and safety standards Requirements
- Tier 1: Bachelor's or Master's degree in a relevant STEM or technical field
- Tier 2: PhD or equivalent advanced industry or professional expertise
- A background in Physics, Nuclear Engineering, Nuclear Physics, Radiological Science, Radiation Physics, Health Physics, Chemistry, Chemical Engineering, Materials Science, or a related engineering discipline
- Professional experience in radiation, nuclear technology, hazardous materials, laboratory science, engineering, or safety may qualify in place of a specific degree
- Based in the US, Canada, the UK, Europe, or Australia
- Analytical judgment and strong attention to detail
- A plus: experience in AI evaluation, AI safety, scientific research, technical review, or quality assurance Expertise
- Remote
- Contract
- Radiological Science
- Health Physics
- AI Safety
- CBRN Benefits
- Fully remote contract work within the eligible locations
- Work that shapes how AI models handle high-risk technical content
- Two tiers, so both first-pass annotators and senior SMEs can take part
- Project-based, on your own schedule