Last June, an attorney filed a legal brief he had written with the help of the generative AI platform, ChatGPT. The document included citations to a series of legal cases that seemingly offered precedents that supported his client’s position. There was only one problem. As the judge in the case discovered, six of those cases didn’t exist. Instead, they’d been dreamed up by the online tool. This was only one of several high-profile incidents in which new technology has stymied — and sometimes embarrassed — the lawyers using it (just Google “I am not a cat.”). Yet, many legal experts believe generative AI will also change the legal profession in ways that will aid lawyers and their clients.
To learn more about how AI is expected to transform the legal profession, as well as other important industry trends in the year ahead, Harvard Law Today recently spoke with Professor David Wilkins ’80, director of the Center on the Legal Profession at Harvard Law School. In the first installation of a two-part interview, Wilkins briefly summarizes legal industry trends he will be following this year and outlines both the opportunities and challenges AI will likely present. In part two, Wilkins will discuss a range of other questions, from lawyering in an election year, to salary trends, to the changing work and priorities of general counsels, and what they mean for the future of legal education.
Harvard Law Today: What do you see as some of the biggest trends in the legal profession this year?
David Wilkins: This has really been a very eventful year for the legal profession, with many important changes and potential disruptions, all of which are interrelated. Let me touch briefly on a few. Leading the list, of course, is artificial intelligence. Every other conversation I have is about ChatGPT and how it will impact the practice of law. Those conversations often lead quickly to talking about legal careers and AI’s potential employment effects. The IMF has said that 40% of all jobs in the world could be affected by AI and that it’s mostly going to be felt in the white collar and professional ranks. That clearly has implications for lawyers.
Another important trend is that, according to 2024 NALP data, the majority of associates at law firms are now women, which has led to a whole series of questions about how diversity will be treated in the legal profession moving forward, particularly in light of the fact that a majority of minority law students are female.
There are also discussions around the role of the legal profession in a time of global unrest. For example, what will be the effects of rising geopolitical tensions — the war in Ukraine, the conflict in the Middle East, competition with China — on markets? Two years ago, ESG [environmental, social, and corporate governance] was the global zeitgeist. And while it is still a powerful organizing force in Europe, it has become a highly contentious issue here in the United States. And, of course, we are entering an election year here in the United States. Lawyers are going to have critical roles to play in helping America and the world protect democracy in a time of increasing polarization and conflict.
Last — but here at HLS, certainly not least — we need to think about what all this means for legal education. How can we best prepare students for the new realities of legal careers in the middle decades of the 21st century? And just as important, how will legal education itself be disrupted by technology and other emerging trends?
“Basic legal information is going to be more and more accessible through technology to more and more people. The problem is that access to basic legal information is just one step in the process of legal services.”
HLT: Let’s start with AI. We’ve seen some high-profile instances of AI gone wrong, when lawyers have used it to write legal briefs, only for the court to discover that the AI made up cases that didn’t exist. What role is AI playing in the legal profession today, where do you see it going in the next year or so, and what checks should we be thinking about putting on its use?
Wilkins: We are just at the very tip of the iceberg in thinking about the implications of AI. When we spoke at this time last year, nobody but perhaps our friends at the Berkman Klein Center had ever heard of generative AI, let alone used ChatGPT. Now, it’s everywhere. In my conversations with lawyers, people started out very skeptical that you could use it for anything useful in the day-to-day work of a practicing lawyer. And now, increasingly, people are more comfortable using it in a wide range of settings.
AI is also getting much better and hallucinating less [a term experts use to refer to inaccurate information generated by AI]. The industry is moving from non-specialized AI to AI trained on legal materials, designed to tackle specific, complex legal problems. Most lawyers I’ve talked to say that if you ask ChatGPT, let alone a more sophisticated version, to write a memo about a legal question, you will get something approximately as good as what a first-year law firm associate would produce. Of course, a lawyer will still have to review it — just as any good senior lawyer will review the work of their juniors before sending it out into the world — but when you think about the relative cost of AI versus a first-year associate, you can begin to see the transformative potential.
HLT: Can you envision a time when clients in need of legal assistance can turn to AI, instead of lawyers, for help, at least in more routine private law matters like divorces or smaller lawsuits? Maybe the plaintiff’s AI talks to the respondent’s AI and to the judge’s AI and the various AIs sort everything out without the humans needing to get too deeply involved. I’m half joking, but then again, did the horse and buggy manufacturers immediately understand the implications of the automobile?
Wilkins: It sounds very futuristic, except we know that things like this are already happening in all sorts of fields. For example, when I get sick, I go to WebMD to see what I might have and what I might be able to do about it, including questions that I should ask my doctor. There are so many sites now where you could get access to information and generative AI is allowing that information to become more interactive. To answer your question, if the issue is access to basic legal information, that is going to be more and more accessible through technology to more and more people. The problem is that access to basic legal information is just one step in the process of legal services. It is already possible to have an AI that fills out basic legal forms—think LegalZoom. It’s clear that the bar is not going to be able to stop this sort of access. More sophisticated, interactive analysis with an AI lawyer? That may now be on the horizon — and more controversial. One of the key challenges in all of this is going to be that access to the kinds of legal tools that AI can provide are not going to be equally available. When I talk to lawyers in the legal services community, they see the potential for AI to help their clients, but they worry that the most sophisticated tools will be in the hands of the most sophisticated, already well-resourced parties who will be able to leverage that technology to gain even further advantage.
“One of the key challenges in all of this is going to be that access to the kinds of legal tools that AI can provide are not going to be equally available.”
HLT: Doesn’t that asymmetry in access to justice exist today? Some people can hire top law firms to help with their legal problems, but many others can barely get access to any lawyer at all.
Wilkins: Of course, you’re absolutely right that some people and organizations have always had more access to legal services than others. As Marc Galanter wrote in his pathbreaking article “Why the ‘haves’ come out ahead,” published half a century ago this year, “repeat players,” including big corporations, who engage in a lot of litigation are always able to invest in a way that gives them an advantage over smaller “one shot” litigants, by, for example, hiring experienced lawyers and playing for the rules by deciding which cases to appeal or investing in lobbying. In one sense, therefore, technology is just another advantage that the privileged will be able to pay for.
But we also know that technology often produces exponential change in ways that create the potential for magnifying existing inequalities exponentially. Thus, it’s not just that some litigants can hire a better lawyer to draft or analyze a contract than others. In the world of AI, these better resourced parties can deploy a tool that can read every relevant contract in ways that no human could ever do. And yet, it is not a simple one-way street, since we have seen how technology also disrupts the powerful. Take the music business. It used to be that the music business was completely controlled by the big record companies, which decided who got to make music and how you bought it — an album, or later CDs (remember those?!) complete with 16 songs and beautiful cover art. And then technology allowed people to order any one song and put them together in any way they wanted. And the record business has been decimated. Now, there are still big players in the record business who figured out how to make money on the top artists. But it’s a whole different business than it was 25 or 30 years ago because of technology. It’s hard to predict exactly how this is going to play out in the legal business, but you are likely to see both an accentuation of the inequality as people at the top are able to access the best technology and therefore get the best results, and the disruption of existing hierarchies as new players enter the legal marketplace offering new kinds of services, leveraging technology in a way that was not possible before.
HLT: I went to school before the age of personal computing, but I’m sure I must have wondered why I needed to learn math when calculators existed to do that work for me. Do we lose something if we become too highly reliant on technology?
Wilkins: Yes, there’s no question. But here’s the challenge. There are two reductive schools of thought, neither of which can be true. One is that every advanced technology is unequivocally good, and the less repetitive labor that humans have to do, the better. The problem with that is that one day we might end up on Barcaloungers like the people in the movie WALL-E. We also know that there’s a huge amount of valuable development that goes into learning the basics, even if once you learned them, you don’t do these elemental tasks anymore. We now let kids use calculators in school once they’ve learned the basics of addition, subtraction, multiplication, and division. Law firms and lawyers need to ask themselves the same question. If we want well trained, mid-level and senior lawyers, those mid-level and senior lawyers need to do some amount of junior lawyer work so that they understand what it is to do in-the-weeds legal research or to craft and write a well-written brief. But the question is, how much of this type of training is needed? If the point is just memorization or information recall, machines will always do it better. Maybe you need to do at least enough to understand what the information is. But you certainly don’t need to do seven or eight years of it just to be a competent partner or a senior level lawyer.