June 2, 2026
Objection: the bot is too good
AI Outperforms Law Professors in Stanford Law Study
Law profs picked AI over each other — and the comments instantly went feral
TLDR: A Stanford study found law professors preferred AI answers to other professors' answers in most blind matchups, suggesting chatbots may be surprisingly good at tutoring law. Commenters split hard between "impressive result" and "dangerous anti-human hype," with bonus drama over whether schools are selling out teachers first.
Stanford dropped a study saying law professors, judging answers blindly, preferred AI-written replies to student contract-law questions over answers written by other professors 75% of the time — and the real courtroom drama immediately erupted in the comments. For some readers, this was a shocking "the robot beat the experts" moment. For others, it was less a breakthrough and more a giant red flag with a Stanford logo on it.
The loudest backlash came from skeptics who saw the study as part science, part power play. One commenter sneered that Stanford seems eager to replace everyone except administrators, calling the whole thing "anti-intellectual nonsense." That instantly triggered its own mini-brawl, with another user swatting the accusation away as flat-out conspiracy talk that "harms discussion." In true internet fashion, the meta-drama got almost as juicy as the study itself.
Meanwhile, a second camp took the results more seriously — if a little nervously. One commenter basically said: sure, 75% is huge, but do we really want machine-written legal help in a system already packed with hidden traps? Another predicted a deliciously ironic future: firms bragging they are "non-AI" the way restaurants brag about organic produce, selling old-school human attention as a luxury product.
The funniest running vibe? A kind of dark comedy: the idea that professors may have just graded a chatbot as the best professor in the room. For a study about legal reasoning, the comments turned into a full-on trial of academia, automation, and who gets replaced first.
Key Points
- •A Stanford Law School-led study tested whether AI could act as an effective tutor for contract law courses using 16 law professors from U.S. law schools.
- •In nearly 3,000 blind, anonymized comparisons, professors preferred AI-generated answers over peer-written answers in 75% of head-to-head matchups.
- •The study used 40 representative contract law questions, with professors writing answers and then evaluating responses without knowing whether they came from AI or other instructors.
- •Professors flagged AI responses as pedagogically harmful 3.5% of the time, compared with 12% for peer-written responses.
- •The researchers also evaluated specific AI systems, including commercial tutoring tools and Google’s NotebookLM, and reported varying performance across systems.