AI Trust + Safety Analyst (French Speakers Wanted to Train AI Models Part Time!)

Role overview

In this fully remote, hourly-paid contractor role, you will evaluate and annotate AI-generated safety content in French and English. Your work will help ensure large language models behave safely across high-risk content categories.

Key responsibilities

• Curate and label safety-focused training examples, including adversarial cases
• Review, score, and compare multiple model responses against safety policies
• Identify failure modes and propose clearer decision rules

Candidate profile

• Near-native or native French proficiency
• Minimum C1 English proficiency
• Experience in Trust & Safety, moderation, or safety evaluation
• Required LLM red-teaming experience
• Strong judgment under ambiguity