AI Red Teamer, CBRNE

📋 Description

• Design technically grounded adversarial evaluation scenarios that test whether AI models
• Evaluate model outputs for technical accuracy, real-world significance, and potential harmful
• Explore dual-use boundaries by testing scenarios that combine legitimate scientific, medical
• Conduct structured multi-step and multi-turn evaluations to assess whether models can be
• Score and categorize model responses using established harm taxonomies, severity frameworks, and
• Document evaluation findings with clear technical reasoning, explaining model strengths

🎯 Requirements

• Graduate-level education or equivalent professional experience in a relevant CBRNE field, such as
• Deep subject-matter expertise enabling accurate assessment of technical content and its potential
• Strong understanding of dual-use research considerations and the distinction between publicly
• Hands-on experience working with multiple large language models, such as ChatGPT, Claude, Gemini
• Strong creative and adversarial problem-solving abilities, with the capacity to think
• Excellent written communication skills, including the ability to explain complex technical risks

🎁 Benefits

• Compensation: $65–$158 per hour, with final compensation determined by factors including location
• Work arrangement: Fully remote within the United States.
• Employment type: Contract position.
• Mission-driven work: Opportunity to contribute to AI safety and help strengthen safeguards around
• Collaborative environment: Work alongside AI researchers, security specialists, policy teams, and
• Professional impact: Apply specialized CBRNE expertise to emerging challenges at the intersection