French AI Data Trainer - Remote

Title: French AI Data Trainer - RemoteLocation: RemoteRole OverviewWe are seeking a detail-oriented and analytically strong AI Safety & Trust Evaluation Specialist to support the development of safer, more reliable AI systems. In this fully remote, hourly contractor role, you will evaluate AI-generated responses, assess reasoning quality, and provide expert safety-focused annotations in both French and English.Your work will directly contribute to improving the safety alignment of large language models by identifying risks, labeling sensitive content, and ensuring outputs comply with strict safety policies. This includes exposure to sensitive material, and requires strong emotional resilience and professional judgment.Key Responsibilities Evaluate AI-generated content for accuracy, safety, and policy complianceAnnotate and label safety-critical datasets in English and French with high linguistic precisionIdentify unsafe outputs related to:Hate and harassmentSexual contentViolence and self-harmBias and misinformationIllegal or harmful activitiesMalicious content or codeCompare multiple AI responses and score them based on safety and quality rubricsIdentify model failure patterns such as evasion, escalation, or unsafe normalizationConduct red-teaming and adversarial testing to uncover model vulnerabilitiesDocument findings clearly, including reasoning behind safety judgmentsSupport refinement of annotation guidelines and edge-case decision frameworksEnsure consistency in policy interpretation across evaluationsRequired QualificationsBachelor’s degree or equivalent experience in Linguistics, Communications, Psychology, Law, Policy, Security Studies, or related fieldNative or near-native French proficiency (reading and writing)Minimum C1 English proficiency (reading and writing)Experience in Trust & Safety, content moderation, policy enforcement, risk operations, or safety evaluationProven LLM red teaming experience (ability to test model safety boundaries and document findings)Strong analytical judgment and ability to work with ambiguous policy situationsHigh attention to detail and ability to explain decisions clearly and consistentlyComfortable working independently in a remote contractor setupPreferred QualificationsPrior experience in AI training data, annotation, or model evaluationFamiliarity with tools such as ChatGPT, Gemini, Perplexity, or similar AI systemsExperience working with structured safety taxonomies or labeling frameworksBackground in investigative or policy-driven analytical roles