Overview
This role involves working with an evaluation team on projects that fine-tune AI models (including large language models). You will apply U.S. intellectual property expertise to review and assess how AI reasons about IP-related legal questions. No prior AI experience is required—experience in U.S. IP law is the focus.
Responsibilities
- Review and evaluate AI-generated responses to U.S. intellectual property hypotheticals, including:
- Patent eligibility (Alice/Mayo)
- Trademark likelihood of confusion
- Copyright fair use analysis
- Trade secret misappropriation
- Use a structured legal rubric to assess:
- Issue spotting
- Accuracy of IP doctrines
- Proper application of the Patent Act, Lanham Act, and Copyright Act
- Identify gaps in AI reasoning around topics such as:
- Claim construction
- Prior art analysis
- Distinctiveness
- DMCA safe harbors
- Provide detailed, high-quality written annotations and feedback to refine AI model outputs
- Collaborate with researchers and project managers to align evaluation standards and project goals
Requirements
- Juris Doctor (J.D.) from an accredited U.S. law school
- Bar admission (active or inactive)
- 3+ years of U.S. legal practice in intellectual property (patents, trademarks, copyrights, or trade secrets)
- Strong reasoning across IP prosecution, litigation, licensing, and enforcement
- Excellent written communication skills and strong attention to detail
- High ethical standards; interest in AI is a plus
Preferred Qualifications
- Patent Bar registration is a plus
Compensation & Benefits
Location
- Flexible, fully remote work arrangement
Commitment & Duration
- Flexible: 10–30 hrs/week
- Duration: ~1 month, with the possibility of extension based on performance and project needs