Law Professors Prefer AI Over Peer Answers

Key Takeaways

• AI answered 40 contract questions, chosen 75% of time
• Gemini 2.5 Pro and NotebookLM matched top professor’s rating
• Study involved professors from 14 U.S. law schools
• Findings suggest AI can provide effective legal tutoring
• Law schools must rethink AI integration in curricula

Pulse Analysis

The legal education landscape is undergoing a rapid transformation as artificial intelligence moves from novelty to necessity. Prior research has demonstrated AI’s ability to pass bar exams, earn top grades, and even grade student work, but the Stanford study adds a new dimension: AI can serve as a peer‑like tutor. By presenting concise, accurate explanations to contract‑law queries, AI platforms such as Gemini 2.5 Pro and NotebookLM are proving they can supplement, and in some cases outperform, traditional faculty guidance.

Methodologically, the study gathered 40 real‑world questions that first‑year contracts students typically ask during office hours. Professors from 14 U.S. law schools authored their own answers, while the two AI models generated parallel responses. In a blind evaluation, the same professors selected the AI‑crafted answers as most beneficial three‑quarters of the time, with the AI matching the highest human rating. This head‑to‑head comparison underscores that AI can handle complex, non‑trivial legal reasoning, not just rote factual recall.

Looking ahead, law schools face a strategic decision: integrate AI tutoring tools into curricula or risk lagging behind peer institutions. Adoption promises scalable, on‑demand assistance for students, potentially narrowing achievement gaps. However, challenges remain, including ensuring model transparency, preventing over‑reliance, and addressing ethical concerns around bias and confidentiality. Institutions that navigate these issues thoughtfully could harness AI to enrich legal education while preserving the critical role of human mentorship.