AI Red Team Tester

About the role

We are seeking curious and creative individuals to join our AI safety team as Red Team Testers. In this fully remote, flexible contract role you will probe cutting‑edge AI models to uncover blind spots and help make them safer for real‑world use.

Key responsibilities

• Design inventive prompts, scenarios and conversational strategies to test AI model weaknesses.
• Attempt to elicit incorrect, unsafe, biased, or otherwise problematic outputs from AI systems.
• Document discovered failure modes with clear, reproducible steps for remediation.
• Assess and rate the severity and potential impact of each issue.
• Collaborate asynchronously with AI safety and research teams to share findings.
• Work across a variety of AI models and task types as new projects arise.

Required profile

• Creative, curious thinker who enjoys puzzles and finding unexpected solutions.
• Comfortable adopting an adversarial mindset and looking for what could go wrong.
• Detail‑oriented and systematic, able to document problems clearly.
• Strong written communicator capable of explaining findings in structured, actionable formats.
• Self‑motivated and reliable, able to work independently without close supervision.

Required skills

What we offer

• Fully remote, flexible schedule – work from anywhere.
• Hourly contract with freelance autonomy.
• Opportunity to contribute directly to high‑impact AI safety projects.
• Potential for ongoing work and contract extensions as new projects launch.