Compliance Law AI Evaluator

Open Train is hiring a Compliance Law AI Evaluator to improve AI systems through expert review of legal and compliance content. You will combine substantive U.S. compliance-law judgment with careful written analysis of accuracy, reasoning, ambiguity, and legal risk.

You will work remotely with project leads and technical specialists on nuanced compliance and regulatory questions.

• Review and annotate legal documents and compliance materials used in AI system development.
• Evaluate AI-generated legal responses for accuracy, reasoning quality, and adherence to U.S. compliance standards.
• Write explanations, rationales, scenario-based questions, and answers for complex compliance situations.
• Identify unclear reasoning, regulatory risks, and potential legal pitfalls in model outputs.
• Suggest precise improvements to AI-generated responses and give feedback on AI training processes.

This is a part‑time contractor role for attorneys who currently practice compliance law in the United States. The role requires recent, substantive compliance‑law experience and strong communication skills.

• Pay: $100-$180 per hour.
• Time: 20 or more hours per week.
• Location:
  
  United States only; remote work.
• Language:
  English.
• Education:
  
  Juris Doctor degree from an accredited U.S. law school.
• License:
  Active attorney license in good standing.
• Experience:
  
  Three to ten or more years of recent, substantive U.S. compliance‑law practice.
• Knowledge:
  Strong understanding of U.S. compliance and regulatory issues.
• Communication:
  Ability to explain complex legal concepts clearly and concisely, in writing and in conversation.
• Helpful background:
  Judicial clerkship experience, teaching law or legal writing at an accredited law school, or legal journal editorial experience.

AI training is the human work behind systems that learn from examples, including reviewing documents, rating model responses, and writing feedback. Experienced professionals are paid to apply their subject‑matter judgment so AI systems produce more accurate, useful, and responsible results.