For a profession that tends to move cautiously, the early response to artificial intelligence was predictably slow. Lawyers debated it. Law schools studied it. Many treated it as something on the horizon rather than something already at the door. But that hesitation has given way to rapid adoption. In a relatively short period of time, AI has moved from curiosity to a basic part of legal practice, both in the classroom and in the day-to-day work of lawyers.

A recent Bloomberg Law survey of ABA-accredited law schools makes that point clear. Twenty-five of the 28 schools surveyed now offer courses focused on AI. Not just one-off electives, but a range of offerings, including introductory classes, advanced courses, and even tech-focused clinics.

The focus is not on turning law students into programmers. It’s on building practical, usable skills. The most commonly taught areas include AI ethics, AI legal research and writing, and prompt engineering. In other words, students are not just being shown what these tools can do. They are being taught how to use them responsibly.

Some schools are going even further. At Washington University, for example, students will participate in a course partially taught by an AI system itself, with the goal of strengthening their ability to critically evaluate AI-generated output. That is a subtle but important shift. The skill is not just using AI. The skill is questioning it. And that may be the most valuable lesson of all.

The survey also reflects a broader trend in legal education. After years of resistance, law schools appear to be moving more toward skills-based training. Law schools now offer significantly more courses focused on practical skills like research, writing, and client work. The long-standing criticism has been that law schools teach students how to think like lawyers but not necessarily how to practice like lawyers. AI is now part of closing that gap.

Not surprisingly, the market is reinforcing this shift. According to the same reporting, many attorneys supervising junior lawyers now expect new hires to be fluent in basic AI-related tasks like cite-checking, reviewing work product, and understanding the ethical implications of AI use. Nearly half expect incoming lawyers to be AI literate on day one.

But here is the uncomfortable part. Most practicing lawyers did not learn any of this in law school. We are being handed tools that can draft, summarize, analyze, and generate at a speed that would have been unthinkable a few years ago. And many of us are teaching ourselves how to use them without any kind of formal training.

That approach might work for learning a new feature in Word. It is not sufficient for something that has the potential to affect competence, confidentiality, and judgment. AI is not just another efficiency tool. It is a system that can be immensely helpful and quietly wrong at the same time. It can produce polished work that looks like legal analysis but includes fabricated citations or subtle errors. It can create the illusion of understanding without the substance to back it up.

And importantly, it does not relieve the lawyer of responsibility.

North Carolina’s 2024 Formal Ethics Opinion 1 makes this point clearly by treating AI, in many respects, like a nonlawyer assistant. The lawyer still has the duty to supervise. The lawyer still has the duty of competence. The lawyer is still responsible for the final work product.

Law schools are beginning to prepare new lawyers to use AI thoughtfully and responsibly. But there is a large group of lawyers already in practice who are expected to use these tools without having been trained to do so. That gap needs to be addressed.

Bar associations, CLE providers, and law firms all have a role to play here. And the focus should not just be on introducing AI, but on teaching lawyers how to use it well.

That means practical training and clear guidance on where AI can add value and where it can create risk. It means talking openly about mistakes—not just the headline-grabbing sanctions cases, but the everyday ways AI can subtly erode quality if it is used carelessly.

There is an opportunity here if we get it right. Used thoughtfully, AI can reduce administrative burden, improve responsiveness, and free lawyers to spend more time on strategy, counseling, and client relationships. But that only happens if the lawyer remains firmly in control of the tool. Law schools are starting to teach that lesson. The rest of the profession will need to catch up. The real issue is not whether AI will become part of legal practice. It already is. The question is whether we take the time to understand it well enough to use it with the kind of care our clients and the State Bar expect. That has always been the standard. The tools have changed. The responsibility has not.