AI Red Teamer

Core responsibilities:

• Red-team products from frontier AI labs to identify security vulnerabilities before deployment
• Perform jailbreaks of advanced models like GPT-5, Claude, and Gemini to uncover potential risks
• Develop Python scripts to systematically test AI system security boundaries
• Work independently with minimal guidance, taking ownership of testing approaches

This role contributes directly to making AI systems safer by identifying and documenting failure modes before deployment, helping prevent potential security risks in advanced AI models.

Required skills and qualifications:

• Experience with AI model testing and security assessment
• Proficiency in Python programming for developing testing scripts
• Ability to work independently with minimal guidance
• Strong analytical skills for identifying and documenting AI system failure modes

No specific benefits, compensation, or perks mentioned in the job description.

Trajectory Labs appears to be an organization focused on AI safety and security, specifically working on testing frontier AI products for vulnerabilities before deployment.