Stanford Law School Professor Julian Nyarko led a study revealing that law professors prefer AI-generated answers to student questions over responses from their peers. The research, involving 16 U.S. law professors and nearly 3,000 comparisons, found AI winning 75% of head-to-head matchups in contract law tutoring scenarios.

The study tested whether large language models could serve as effective tutors, focusing on judgment-rich legal reasoning rather than simple factual recall. Professors evaluated anonymized responses without knowing their source, rating AI answers significantly higher while flagging them as pedagogically harmful only 3.5% of the time versus 12% for peer-written answers.

Co-authors from Yale, NYU, and University of Chicago emphasized that legal education demands nuanced analysis where multiple arguments can be valid. The research team calibrated AI responses to match human answer length and structure, using multiple evaluation methods to ensure validity.

These findings arrive as law schools nationwide debate AI integration. While Nyarko cautions against wholesale adoption, the data suggests blanket skepticism may be unwarranted. The conversation should shift from whether AI can provide quality responses to how it can be deployed responsibly to benefit student learning.