Anthropic’s brand new “Claude for Legal” suite of AI tools for law and lawyers accelerates – dramatically – the pace of AI technology deployment in and around law schools. Seth Chandler (University of Houston) has a characteristically thorough (and completely geeked out) review of how law students and law professors (especially clinicians) will find abundant uses for these systems.
And yet.
As Professor Chandler’s detailed review makes abundantly clear, all of the capabilities that Claude for Legal offers to law students, law teachers, and even to practicing lawyers … are fast-paced, thought- and difficulty-removing versions of exactly the same cognitive and hands-on “lawyering” skills that US legal education has been promoting and building on for decades. There is absolutely nothing revolutionary or even evolutionary here; it is “same old law school, but with many of the hard parts removed.” All of the Socratic rigor of Professor Kingsfield and none of the inefficiency or inhumanity! And none of the life lessons that were the real takeaways from The Paper Chase, the book as well as the film.
Oh, and “keep up with the law school Joneses” – always a tried-and-true leadership strategy for Deans and Provosts – likely now means (if it did not already mean) hiring a tier of legal tech staff experts to steer both faculty and students in productive directions with these systems. Even for AI enthusiasts (whose numbers may grow, if more slowly on law faculties than among students), the learning curve for this stuff is steep.
Exactly what does progress look like (and cost)?
As some readers know, for some time off and on I have plugged away at prompting the profession and others – lawyers, law professors, Deans, technologists, and even leaders in higher education – to grapple genuinely with how legal education systems of the future might (and in my view should) look and act quite different (differently?) in the 21st century. A system that co-evolved 100 years ago with large law firm bureaucracies has little obvious utility today; it persists largely because of inertia, path dependence, and the rewards of an early 21st century status hierarchy that resembles little so much as a late 19th century status hierarchy.
Claude for Legal doubles down on that system, promising to do little more than automate big chunks of it (whether that will happen in practice … we will see). We should never have expected otherwise. Nevertheless, the resources and energy that have been and will be directed at it by legal educators (both embracing it and rejecting it) would be more productively directed to the difficult task of imagining and building a system that is genuinely, broadly, and above all prospectively valuable.
