Ad: BlueJ Better Tax Answers. -Accomplish hours of research in seconds -Instantly draft high-quality communications -Verify answers using a library of trusted tax content. Learn more

Schultz and Fromer on Pathways for Incorporating AI into Legal Education

Jason Schultz and Jeanne Fromer (both NYU Law) have posted a new manuscript at SSRN, titled “Responsible Pathways for Incorporating AI into Legal Education.”

2The abstract:

Generative AI is already reshaping legal practice, and law schools that ignore it or ban it outright risk graduating students who are unprepared to use it and unable to supervise it. We propose a middle path built around two empirically backed observations. First, to use AI effectively and ethically and to mitigate AI’s cognitive risks, students must build sufficient expertise to supervise AI systems the same way lawyers supervise non-lawyer colleagues in practice. Second, because AI is rapidly displacing much of the traditional legal work used to train junior lawyers, law schools must prepare graduates to enter the legal workforce at a more advanced level than in the past, including with the ability to supervise AI as part of their practice.

Building on these observations, we recommend that law schools (1) allow student use of AI on course assessments only after students have demonstrated baseline expertise without it, most fundamentally in the 1L year, though also across the entire law school curriculum; (2) permit AI as a study aid-along the lines of a “tutor” model-outside the classroom but not during class participation; (3) label every course assessment as either human-only-taken without devices or in secure exam software-or AI-assisted-in which students have the option to use AI if they also document and describe how they supervised the tools they used; (4) encourage lawyering programs, clinics, externships, labs, and simulation courses as well as pertinent cocurriculars to become a specialized training ground for ethical AI supervision where students can engage with the opportunities and risks of AI in situ under the supervision of practicing faculty; and (5) implement this framework through a standing committee, a formal learning outcome under ABA Standard 302, and a one-year pilot that is evaluated before any school-wide adoption. That committee could also review the possible risks and benefits of new uses of AI in teaching and assessment.

As we hope these recommendations make clear, we think that only faculty who are interested in using AI in their courses and assessments should do so, and those who are not should not feel pressured to. Still, as AI becomes part of practice, each instructor should at least make a deliberate choice, weighing the learning value to students (and, in clinics, to current and future clients) against any ethical or pedagogical objections, rather than leaving the question unaddressed. A few proposals, such as labeling each assessment and setting a default rule when a syllabus is silent, would apply to every course; the rest are offered as considerations for faculty who choose to engage with AI.


About the Author

Ad: BlueJ Better Tax Answers. Blue J's generative AI tax research solution is transforming how tax experts work. Learn more.
Information and rates on advertising on TaxProf Blog