The Non-Tech Law Professor’s Guide to Building an AI Law Course

Your students are already using AI in their work. Are you ready to teach them how to do it responsibly?

If that question made you uneasy, you're not alone. In conversations with law faculty across the country, the same hesitation comes up: I'm not a technologist. I don't know how these systems work. How am I supposed to teach this?

Here's what I want you to hear: you don't need to be a technologist. You never did. The attorneys who will face the highest-stakes AI decisions in the next decade won't be the ones who can explain a neural network. They'll be the ones who know how to question it, govern it, and hold it to a professional standard.

That's not computer science. That's lawyering. And you already know how to teach that.

This article is a practical guide for law educators who want to build an AI law module or course — not from scratch, and not alone. By the end, you'll have a content framework grounded in what law firms actually expect, a clear sense of which topics you already own, and a faster way to get it into your students' hands than you thought possible.

The AI revolution is real — and your students can't afford for it not to be in the curriculum

Let's start with an honest assessment, because your students deserve one.

The noise around artificial intelligence in law is enormous, and much of it is hype. AI will not replace lawyers overnight. It will not autonomously argue cases or render judgments. The gap between what AI vendors promise and what their tools reliably deliver is still significant — and closing that gap requires exactly the kind of critical, evidence-based thinking that legal training produces.

But the revolution is also real. The distinction that matters isn't between AI and no AI. It's between AI as an assistant — a tool that helps a lawyer do something faster — and AI as an agent: a system that initiates, executes, and completes tasks with limited human direction. That shift is already underway inside law firms. Document review, contract analysis, litigation prediction, knowledge management — these are not future applications. They are current ones, running now in firms where your students will work in three years.

The graduates who walk in without any framework for evaluating, questioning, or governing these systems will not be incompetent. But they will be behind. And in a profession where preparation is the baseline expectation, behind is a risk neither they nor their clients can afford.

This is not a reason for alarm. It is a once-in-a-generation curriculum opportunity — and the fact that you're reading this means you already understand that.

Start where law firms already are: the competencies they expect your graduates to have

The most common mistake in building an AI law course is starting with AI and working toward law. Start instead where your students are headed: the firm, the courtroom, the regulatory agency — and ask what those institutions already need.

The answer is clearer than you might expect. Law firms that have implemented AI training programs have converged on a shared set of required competencies: the ability to evaluate whether an AI tool is appropriate for a given task; the ability to supervise AI-generated output with the same rigor applied to junior associate work; an understanding of where AI introduces legal and ethical risk; and the capacity to communicate AI use — and its limitations — to clients transparently.

There is also a harder truth worth naming directly: many law firms are now building internal AI training programs because they don't trust that law schools are doing it. That is not a criticism — it is an opening. Law schools that build rigorous AI curricula now will produce graduates who arrive prepared rather than requiring remediation. That matters to hiring partners. It will matter to your students' careers.

Your course doesn't need to cover everything. It needs to cover what firms are measuring. Here is a practical starting framework:

• Competency 1: Evaluate AI tool suitability for a specific legal task
• Competency 2: Supervise and verify AI-generated legal work product
• Competency 3: Identify AI-related ethical obligations under the Model Rules
• Competency 4: Assess and communicate AI risk to clients and colleagues
• Competency 5: Apply due diligence standards to AI vendors and platforms

Five competencies. Five modules. That is a course.

Module one belongs to ethics — and the bar already wrote most of it for you

If you have taught professional responsibility, you already have the foundation for the strongest module in your AI course. Artificial intelligence did not create new ethical obligations for lawyers. It created new fact patterns that trigger the ones that already exist.

Competence now includes technological competence — the obligation to understand, at a sufficient level, the tools being used in legal practice. Confidentiality requires that attorneys understand where client data goes when it is submitted to an AI system, and who has access to it. Supervision means that AI-generated work product must be reviewed with the same diligence applied to any work done by someone working under your direction. The duty of candor extends to AI-assisted filings — a hallucinated citation submitted to a court is not the AI's problem. It is yours.

These are not hypothetical concerns. Courts have already sanctioned attorneys for submitting AI-generated briefs that cited nonexistent cases. Bar associations across the country — including the ABA, whose partnership with LiveCarta makes these materials available — are issuing guidance at an accelerating pace. The frameworks in that guidance map directly onto the professional responsibility doctrine your students already study.

If anything, AI makes professional responsibility more urgent, more interesting, and more teachable — because the stakes are visible, the cases are new, and the doctrine is still being written.

Risk literacy: the module that separates a good AI course from a great one

Ethics tells students what the rules are. Risk literacy teaches them what goes wrong when the rules meet reality — and that is where the most valuable learning happens.

The risk landscape in AI-assisted legal practice clusters around five areas, each of which maps cleanly to a seminar session or case study: data privacy and security; confidentiality breaches from AI vendor data practices; bias and discrimination embedded in AI outputs; regulatory compliance across jurisdictions; and malpractice exposure when AI error causes client harm.

Consider the malpractice scenario as a teaching exercise. An attorney relies on an AI-generated case summary to advise a client on litigation strategy. The summary contains a hallucinated precedent — a case that does not exist, cited with full case name, volume, and page number. The attorney does not verify it. The advice is wrong. Walk your students through the liability analysis: competence, supervision, causation, harm. Every element of that analysis is doctrinal. The AI is simply the instrument through which a familiar failure occurred in an unfamiliar way.

The vendor due diligence component is equally important, and equally underserved. Students entering practice will be asked — formally or informally — to evaluate legal technology vendors. They need a framework: What data does this tool access? Where is it stored? Who trains on it? What happens to client information after a matter closes? What contractual protections exist?

How to actually build it: from blank syllabus to shareable course in days, not months

Here is where many faculty get stuck — not on content, but on assembly. You can know exactly what you want to teach and still spend months trying to find materials that fit together, cover the right topics without overlap, and land within a budget students can actually afford.

This is the problem that LiveCarta was built to solve — and specifically why the ABA partnership makes these five AI law titles available through the platform.

Here is what the build process looks like in practice.

Step one: generate your outline. The LiveCarta AI Curriculum Builder lets you type a topic — "AI ethics in legal practice," "artificial intelligence and professional responsibility," or simply "AI law" — and generates an 8-week course outline with module suggested readings drawn from the catalog. What would have taken several weeks of syllabus drafting takes about ten minutes. You can adjust the number of modules, review the suggested chapters, and add items to your cart directly from the outline.

Step two: build your Collection. Rather than assigning one textbook that covers AI law at 60% relevance, you select the chapters that match your modules precisely. Week one gets the AI revolution overview. The ethics and risk unit gets two chapters that cover exactly those topics — and nothing else. Students pay only for what they read, which means lower costs and higher relevance.

Step three: personalize material. Once you have selected your chapters, you can annotate directly within the material. Add a note before the malpractice section flagging the hypothetical you'll use in seminar. Embed a video of an ABA panel on AI governance. Reorder the sequence to match your pedagogy. The book becomes your book, without requiring you to write one.

Step four: assign it. Compile your Custom Copy, your Collection, and any supplemental documents — ABA guidance, law review articles, case studies — into a single teaching material. Students click one link and have everything they need from day one. If you have a co-instructor or teaching assistant, they can update the Custom Copy, Collection, or Bundle mid-semester without redistributing anything.

The entire process can be completed in a single focused working session. That is not a marketing claim. It is a design choice: the platform was built for educators who have a full course load and cannot spend a semester building one course.

What you're really building: the future-ready lawyer your students need to become

It is worth stepping back from the logistics for a moment to name what is actually at stake here.

The legal profession is not going to be replaced by AI. But it is going to be reorganized by it. The roles that will define the next decade of legal practice — AI governance counsel, legal technology evaluator, algorithmic risk advisor — do not all have formal titles yet. But they are already taking shape inside firms that are making strategic AI decisions right now.

That is not a small thing to teach. It is, arguably, the most consequential curriculum decision your department will make this decade.

The professor who builds this course is not updating a syllabus. They are deciding what it means to be a lawyer in 2030 — and giving a cohort of students the framework to meet that moment with competence, with ethics, and with confidence.

The materials exist. The tools to build with them exist. The students are waiting.

The five AI law titles — including The AI-Enabled Law Firm — are available by individual chapter on LiveCarta, made possible through LiveCarta's partnership with the American Bar Association. Use the AI Curriculum Builder to generate your first course outline, or explore the catalog to start building your Collection.