Nuclear Safety AI Model Auditor & Risk Assessor

AI Trainer Jobs; United States; Remote; USD 74,000 - 124,000; Full time

Job summary

AI Trainer Jobs is seeking specialists to audit AI model responses to sensitive nuclear questions. You will classify requests as benign, dual-use or adversarial, and evaluate model outputs for safety and accuracy.

Your task includes ranking three models and drafting the ideal responses, contributing to safer, higher-quality AI systems. Open to qualified nuclear engineers, radiological scientists, and safety professionals with strong analytical and written communication skills.

Qualifications

• Nuclear engineering or health physics background with at least five years of professional experience.
• Experience auditing AI model outputs and safety classifications.
• Ability to evaluate dual-use and adversarial requests.

Responsibilities

• Audit how AI models respond to sensitive nuclear questions.
• Classify user requests as benign, dual-use or adversarial and justify the classification.
• Mark each model's answers as safe or unsafe and explain why.
• Rank the three models and draft the answer the model should have given.

Skills

Nuclear Physics

English

Physics reasoning

AI Training

Scientific review

Nuclear Structure

Fusion

Fission

Radioactive Decay

Radiation

Education

Relevant degree

Job description

AI Trainer Jobs is seeking specialists to audit AI model responses to sensitive nuclear questions. You will classify requests as benign, dual-use or adversarial, and evaluate model outputs for safety and accuracy.

Your task includes ranking three models and drafting the ideal responses, contributing to safer, higher-quality AI systems. Open to qualified nuclear engineers, radiological scientists, and safety professionals with strong analytical and written communication skills.