Responsibilities
• Curate and label safety-focused training examples, including adversarial red-team cases
• Review, score, and compare model responses against safety policy rubrics
• Document why outputs are safe or unsafe and identify failure modes
• Stress-test and audit model behavior for policy gaps and edge cases

Requirements
• Native or near-native French proficiency (reading/writing)
• Minimum C1 English proficiency
• Experience in Trust & Safety, content moderation, or policy enforcement
• LLM red teaming experience required
• Strong knowledge of safety domains (hate, harassment, self-harm, violence, bias)
• Emotional resilience for explicit content exposure

Preferred
• Previous AI data training or annotation experience
• Experience with Perplexity, Gemini, ChatGPT, and similar tools

How to Apply
Click "Apply" to be taken to the SME Careers website. The application process involves filling out your profile, completing an interview, and taking an exam for certain roles. If you pass, you will receive an email and your profile will be reviewed before being contacted for project onboarding. If you do not pass, you can redo the interview. Payments are made weekly once work is approved.