Beth Williams’s desk is buried in vendor proposals, pilot agreements, and NDAs. Virtually every week, she says, brings another AI agent or platform promising to redefine how legal research is done.

“We’re reviewing more products at once than I’ve seen in my career,” says Williams, the Robert Crown Law Library’s associate dean and a senior lecturer. “It is probably a bit like the early days of Lexis and Westlaw, but this is that moment on steroids.”

AI tools arrive in a steady stream of bright, shiny releases, but they account for only a small part of what the Stanford Law librarians are managing these days. Over the past two and a half years, Williams and her team have developed an AI framework that stretches across research guidance, pedagogy, technical development, and training for the full law school community. To extend that work, the library recently added two AI-focused librarians—one specializing in data science and AI, the other in emerging AI technologies—positions Williams believes are the first of their kind in U.S. academic law libraries.

“Among other things, these roles are allowing us to build more tools ourselves, not just consume what the market produces,” she says. “We have been able to move forward with a lot of momentum for a few reasons. We have the advantage of geography, of course, but it’s more than that. Our librarians were experimenting with AI long before anyone asked them to. In fact, much of our AI infrastructure happened organically. People on the team saw where the future was headed and moved quickly in that direction.” 

An initial catalyst was Lexis’s rollout of its first AI tool in 2023, says Taryn Marks, associate director of research and instructional services. She and Will Huggins, associate director for library technology and innovation, began testing it, along with tools from Westlaw and Bloomberg, so the material could be incorporated into Marks’ next legal research class. 

“We spent weeks figuring out what each tool could actually do,” Marks said. “And we started building guidance around those differences because our students and faculty needed clarity.” 

They built an internal site that captured the team’s tests, tool comparisons, and step-by-step workflows. Soon Huggins started circulating a weekly AI update summarizing major model changes and vendor developments. And they rolled out an ever-expanding set of workshops for students, faculty and staff, from modules on building writing partners to how-tos on using NotebookLM for document analysis. Recordings and materials live on the library’s AI Learning Hub, a resource that houses past sessions, upcoming trainings, and practical guides.

And anyone who needs help can drop by the Curiosity Corner for one-on-one assistance on everything from prompt-writing to result evaluation.

All of this has attracted attention beyond Stanford as other law schools now request their syllabi, a sign, Williams says, that peers view Stanford as a model in this space.

The library recently launched a one-credit AI Literacy for Lawyers course that introduces students to the fundamentals of working with AI in legal practice: how to evaluate outputs, understand risks, and act as critical, ethical consumers of these tools. “Our students have been using these tools for more than two years,” Williams says. “The course gives them a structured way to understand what they’re actually doing.”

Meanwhile, Huggins has been developing apps to help them do more. His oral argument practice app offers students targeted feedback, and he helps students and others at the law school build similar tools. He even created an AI-supported scheduling system for the library’s service desk, an experiment that has since become part of the library’s everyday operations—underscoring how AI is now woven into the way the library functions as much as how it teaches.

Learn More About the Library’s AI Initiatives