AI Policy Reviewer – Remote

About the role

We are looking for an AI Policy Reviewer to evaluate AI‑generated and user‑generated content, ensuring it meets our governance standards, regulatory requirements and responsible AI principles. Working remotely, you will act as a quality‑assurance gatekeeper, protecting model integrity and user safety.

Key responsibilities

• Review and score AI‑generated responses against detailed policy rubrics, assessing safety, truthfulness, fairness and alignment with community guidelines.
• Identify and flag instances where the model is over‑ or under‑sensitive, handling ambiguous edge cases and making nuanced judgement calls.
• Analyze data to uncover systematic flaws such as bias, hallucinations or policy gaps, and report findings to Product and Engineering teams.
• Collaborate with Policy teams to test and refine evaluation rubrics and provide feedback on teachability of policies.
• Participate in adversarial testing (red‑team) to attempt jailbreaks and expose vulnerabilities before launch.
• Work closely with Machine Learning Engineers, explaining rating decisions and helping adjust model behaviour.
• Create high‑quality prompt‑response examples that serve as “golden sets” for training.

Required profile

• Minimum 3 years of professional experience in Trust & Safety, Content Policy, Risk Analysis or Legal/Compliance review.
• Deep knowledge of content moderation policies covering hate speech, harassment, misinformation and graphic violence.
• Strong analytical ability to deconstruct AI responses, detect logical flaws, hallucinations and subtle biases.
• Excellent written communication to explain findings to engineers and policy experts.
• Emotional resilience and self‑care strategies for exposure to disturbing content.
• Familiarity with large language models such as ChatGPT or Gemini.
• Comfort with dashboards, spreadsheets and specialized review tools.
• Awareness of global cultural and political nuances.

Required skills

• Experience with LLM platforms (e.g., ChatGPT, Gemini)
• Proficiency in content‑moderation review tools
• Ability to work with dashboards and spreadsheet analysis