A Con Law Professor & Law Librarian Enter the Breach: Teaching AI-Integrated Legal Analysis
As the initial wave of enthusiasm surrounding AI recedes, it gives way to diminishing returns and a field shaped by the diverging perspectives of its most influential thinkers. As Gen AI tools continue to emerge in rushed, disjointed, and deeply imperfect forms, law schools find themselves in a reactive posture, scrambling to mitigate the shortcomings of technologies they never had the luxury of adopting on their own terms. Once more, law librarians must fill the breach: evaluating tools, designing ethical workflows, and training others to navigate the uncertainty.
A January 12, 2026, ABA Journal article found that many academic law librarians have integrated AI into legal research instruction and launched initiatives to prepare students who already expect to use it. Law schools have created AI-focused librarian roles, reflecting a rising demand for their expertise. Law librarians now serve as key advisers on faculty committees, guiding how AI can support scholarship and contribute to the broader professional conversation about responsible adoption. As one law school director of AI and legal technology put it:
Do you want a random con law professor teaching AI? Or do you want somebody who has a background in information science teaching AI?
Introducing AI-Integrated Legal Analysis
In January, my law school introduced what may be a first-of-its-kind course: AI-Integrated Legal Analysis. Taught by a not-at-all-random con law professor and a loud law librarian in collaboration, the course equips students with the critical skills to responsibly and effectively incorporate AI tools into legal research, writing, and decision-making. Rather than relying on abstract theory, students work through hands-on exercises and case studies in a true legal laboratory setting, using platforms like Claude, Gemini, Protégé, and CoCounsel (and more) to evaluate AI outputs, detect hallucinations, and maintain professional standards.
The syllabus treats AI as an unsettled, rapidly moving technology—one that no one fully understands how to teach—positioning the course as a collective experiment in responsible legal practice under uncertainty. At the outset, the very specific con law professor and loquacious law librarian eagerly admitted that they expect to learn as much (if not more) than the students. Each week blends doctrinal readings, empirical research, regulatory materials, and hands-on exercises, producing an iterative pedagogical cycle: read, use, evaluate, correct, and reflect.
Course Content
The readings deliberately expose students to the full ideological spectrum of AI worldviews. At one end is doomerism—materials that echo Eliezer Yudkowsky’s existential-risk warnings, as well as documents such as the “Right to Warn” and the 2025 Superintelligence Statement. In the middle is the moderate pragmatism of scholars like Ethan Mollick, who sees AI as transformative but error-prone and manageable with proper human oversight. At the other extreme is the accelerationist camp (including a slew of random tech bro billionaires)—Yann LeCun, Sam Altman, Jensen Huang, Elon Musk, and others—who argue that rapid scaling will yield unprecedented capability gains and should not be slowed by speculative fears.
A distinctive feature of the course is its curated “Pick?One” media list for Week 1, which asks each student to select and bring into discussion a single film, episode, interview, or documentary that probes different cultural framings of AI. The options span more than five decades of public imagination—from the 1970 techno?paranoia of Colossus: The Forbin Project to cyber?anxiety episodes of The X?Files, to contemporary analyses such as MIT’s AI Snake Oil lecture and the documentary podcast series Geoffrey Hinton vs. The End of the World. Other offerings include long?form discussions on topics such as the AI apocalypse risk, the socioeconomic “rot economy” surrounding tech platforms, and selections spotlighting the perspectives of AI doomers, scouts, and accelerationists. This diverse media sampling ensures the seminar begins with a kaleidoscope of intellectual perspectives and emotional tones, linking the course’s technical and legal inquiries to the broader cultural narratives shaping public understanding of AI.
Practical Applications & Outcomes
Against this backdrop, the course’s problem sets function as structured experiments in model behavior. Students run multiple chatbots through classic legal tasks (conflict of laws, contracts, arbitration clauses, right of publicity, and motion practice), then diagnose hallucinations, reasoning failures, and accuracy gaps. Critically, the problems require students not only to generate outputs but also to decide which follow-up questions a competent lawyer would ask, reinforcing that AI is a tool that requires supervision, not deference. Additionally, the course will collect a small sample set of data on chatbot performance.
The course’s distinctive contribution is methodological: it treats AI not as a settled doctrine but as a contested frontier. Students are forced to confront flawed, fast-moving technologies while navigating the competing intellectual frames, from existential dread to cautious optimism to aggressive acceleration, that will shape the world in which these tools operate. Importantly, the course instills in students the need to add value beyond what the AI produces: synthesizing outputs into coherent recommendations, identifying weaknesses and ambiguities, interpreting holdings rather than reciting them, and integrating the specific facts of each case. In practice, this also means knowing when independent research is still required.
Next Steps
Long ago, Joy Division’s Ian Curtis hauntingly crooned a question that feels newly relevant: So what ya gonna do when the novelty has gone? The AI boom has already begun its contraction. The breathless headlines are quieter. The promises are hedging. The tools that were supposed to transform everything have proven powerful, unreliable, and deeply dependent on the humans who use them. AI visionaries who slapped each other’s backs and pretended they knew all along what AI would mean for legal practice often find themselves without useful answers.
The honest answer is that no one knows what comes next. The ideological spectrum the course maps, from existential dread to reckless acceleration, reflects genuine uncertainty, not a debate waiting to be resolved. What we do know is that the attorneys who survive this moment will not be the ones who deferred to the technology or dismissed it. They will be the ones who kept working, kept questioning, and resisted the comfort of belonging to any particular camp. The course does not promise a destination. It promises the habit of mind that makes the uncertainty navigable. Finally, this blog will end by twisting Joy Division’s Novelty. File fake briefs while you can, but don’t ever relax, because sanctions lurk for those who slack. The foregoing atrocity exhibition was not AI slop, but this blogger’s unique blend of hermetically sealed mordant humor.
