This weekend, I returned home from the fabulous AALS ADR Section Works in Progress conference at the University of Missouri. (Not a long trip for me.)
As always, I really enjoyed the presentations and conversations with friends and colleagues. By all accounts, Carli Conklin and Rachel Wechsler did an outstanding job planning and hosting the conference.
After I presented my short article, Beyond AI Slop: Preparing Law Students for Responsible Practice in the AI Era, an associate dean asked for my thoughts about developing an AI strategy for her school. We didn’t have enough time to discuss it, and I thought that others might be interested. Many members of our community are current or former – and probably future – deans and associate deans. Faculty who never become administrators also will help establish school policies through committees and other activities – and will implement those policies in their courses and work with students.
Why Every Law School Needs an AI Strategy
Today’s law students will graduate into a world where AI is increasingly part of lawyers’ regular work – but that’s only part of the challenge. They will receive AI-influenced material from counterpart lawyers, clients, self-represented parties, experts, and others. Some of it will contain hallucinations. Some will be inaccurate, misleading, poorly analyzed, excessive, or otherwise unreliable. Clients may develop unrealistic expectations from their own AI use, disclose confidential or sensitive information, or create AI chats containing “smoking gun” material that may turn out to be discoverable.
Lawyers therefore need to know how to use AI responsibly, evaluate potentially unreliable material from others, and counsel clients about their own AI use. Law schools that don’t prepare students for these tasks risk graduating students who produce AI slop themselves or don’t recognize and deal with it effectively when others do. Poorly prepared lawyers face greater risks of financial and disciplinary sanctions for filing unreliable AI-generated material, avoidable litigation problems resulting from clients’ AI use, and difficulties counseling clients whose unrealistic expectations have been shaped by AI. These problems won’t arise in every matter, but lawyers need to be prepared to deal with them in any matter from their first day on the job.
What Should a School-Wide AI Strategy Include?
Law schools should carefully plan how students will develop competence in dealing with AI. Each school should develop an effective strategy appropriate to its circumstances. Each strategy should extend across students’ entire educational experience rather than relying on isolated electives or individual faculty policies. Schools can use orientation programs, workshops, legal research and writing programs, clinics, simulations, specialized courses, library programs, and other activities.
Schools don’t all need the same strategy. A good strategy at one school may not work well at another. Schools have different students, faculties, curricula, resources, and educational goals. A coordinated strategy doesn’t require every professor to teach AI or use it in every course.
The Beyond AI Slop article suggests several principles that law schools can adapt to their own circumstances:
Establish a minimum standard of competence for every graduate. Every student should acquire basic competence in responsible AI use rather than leaving AI education entirely to electives – or worse, leaving students to learn about AI on their own.
Develop both AI and non-AI competence. Students need strong independent legal research, analysis, writing, and judgment so they can recognize when AI produces unreliable work. Schools should consider how to sequence students’ development of independent legal competence and their use of AI so that students can evaluate AI-generated work effectively.
Teach verification as a professional skill. Students should develop systematic, risk-based habits for checking important information, whether generated by AI they use or received from someone else.
Give students experiential practice and assess their performance. Students should actually use AI, diagnose problematic outputs, correct them, and make judgments about whether, when, and how it should be used. Schools should assess whether students can perform these tasks competently.
Include AI education as part of professional responsibility and client counseling. Students need to consider confidentiality, ethics, client communication, decision-making, and their responsibility for work produced with AI assistance.
Create institutional responsibility and periodically reassess the strategy. AI and legal practice are changing quickly, so schools should designate individuals or a committee to periodically reconsider what students need.
Law Schools’ Role in a Larger Public Legal-Health Strategy
AI education in law schools should be part of a larger public legal-health strategy. The public legal-health concept draws on the preventive orientation of public health, emphasizing practical ways to prevent problems and reduce harms that cannot realistically be eliminated.
In my blog post, Bummed About AI? Democratic Societies Need Public Participation to Shape It, I discuss a combination of strategies, including regulation, industry practices, political and social pressure, professional norms, and education. These strategies can reinforce each other. Education is especially important because people need to understand AI well enough to use it wisely, question it appropriately, and participate effectively in decisions about its development and use.
This connects with discussions at the conference about democracy. Our field can contribute to a healthier democracy by helping people understand difficult situations, consider different perspectives and tradeoffs, evaluate realistic options, and make thoughtful decisions under conditions of uncertainty.
Those capacities are increasingly important as AI raises major public policy issues. Recent reports about AI systems autonomously hacking into computer systems are one dramatic example. Questions about privacy, surveillance, misinformation, employment, regulation, concentration of power, and many other issues will require important public choices. Our society will be better able to make those choices with an informed and engaged public.
Lawyers have a particularly important role in this process. They will help clients, institutions, governments, and communities make decisions about AI, and they will participate in developing the laws, policies, professional norms, and institutional practices that shape its use.
Preparing law students to understand AI therefore can serve two related goals: preparing them to practice responsibly and to participate thoughtfully in the democratic decisions that will shape AI’s role in our society.
As you may have guessed, I directed RPS Coach to help me express my views on these issues as effectively as possible.
