ChatGPT think again

Think ChatGPT will eliminate lawyers? Think again.

ChatGPT, the chatbot developed by OpenAI, has set the legal industry buzzing with speculation about how this powerful AI will impact the profession. 

Many are saying lawyers shouldn’t use AI, that it’s dangerous. Or that the tech might eliminate the need for lawyers. Some are even outright advising lawyers to “be afraid,” claiming it will make them obsolete. 

But these fears are unfounded. In fact, the opposite is true—AI will be a boon to the profession. Let’s take a moment and shed a little light on some of the most prevalent “predictions.” 

Fear 1: AI will make lawyers unnecessary

First, let’s recognize that today more than 80% of Americans can’t afford a lawyer. And more than 90% of the legal needs of low-income Americans go unmet. Clearly what we need is not fewer lawyers, but more affordable ones. 

So imagine a day when a lawyer has access to an AI tool built on a large language model like that behind ChatGPT—a product that’s been trained and vetted by experts in legal tech to truly be able to handle and be trusted with substantive legal tasks. That lawyer might delegate several tasks to this specialized AI—drafting letters to opposing counsel and clients, reviewing and summarizing documents, perhaps even performing some legal analysis and research. 

What’s left for the lawyer to do? That’s easy: The dozens and dozens of things AI can’t do—make strategic decisions, counsel clients, attend court appearances, and take depositions, to name just a few. In fact, these are among the many things most any lawyer will say they wish they could prioritize, but simply don’t have time for.

Lawyers won’t be replaced—they’ll be working on problems AI can’t solve. While AI can in just minutes review thousands of documents or draft a contract clause, it can’t make decisions about what to do with that output and how it matters to a case. It can’t decide to ask the question in the first place.

The shift in the industry will go something like this: at first the cost of legal services will decrease, while the value of those services increases. AI will take care of the most time-consuming work, giving lawyers the time to put more effort into strategy and intelligence—human work. A more valuable, more affordable product means more of the people who can’t afford a lawyer today can. This increased demand ultimately leads to more jobs—which is precisely what happened when ATMs were introduced to banking

Reality check: Lawyers will be needed more than ever, and they’ll be spending their time on more valuable, more rewarding work than they do today. 

Fear 2: AI will kill billable hours 

Talk of “the end of the billable hour” has been going on for decades—AI is just the latest “culprit.” 

To claim AI won’t impact legal billing would be inaccurate, and disingenuous. But the big shift won’t be the death of the billable hour—it’ll be how a lawyer’s work is valued. As more legal work is competently handled by machines, the more time and focus a lawyer can dedicate to the kinds of work that have always been the most valuable—and that machines can never replace. These include knowing what questions to ask, developing legal strategy, issue-spotting, and applying expertise to complex legal questions. 

Consider document review: It’s tedious, but it has to be done. But young associates today spend a significant portion of their document review time simply identifying and locating the relevant facts in the first place. Using AI to do the “first cut” quickly—in minutes, not hours or days—propels lawyers past the “slog” and right into developing and applying their critical thinking and creativity to the relevant facts surfaced by the AI. 

After months of working this way, the average young associate will have spent more time on this more strategic and challenging work—for more cases—to their own, their firm’s, and most importantly, their clients’ benefit. Using AI, lawyers will work just as many hours, perhaps even more (see reality check 1), but they’ll spend those hours on much more valuable work, for more people.

Reality check: AI doesn’t mean fewer hours of work. It means more of that time spent doing work that builds lawyers’ expertise and increases their value.

Fear 3: Because AI’s output is sometimes inaccurate, lawyers shouldn’t use it

ChatGPT is making headlines both for its impressive capabilities and its flawed output. In some instances, the chatbot produces “hallucinations,” seemingly correct answers that are actually inaccurate. And ChatGPT—like all AI—demonstrates some troubling bias. Obviously, lawyers should use AI tools with full awareness of their strengths and limitations, and not rely on them to think for them. Like any work done by, say, an assistant, AI output should be carefully reviewed for accuracy and appropriateness.

AI will only continue to improve, and that’s good news. But even more important to remember is that while a machine like ChatGPT understands and interacts with language in a truly revolutionary way, it’s still quite limited. That’s because it can only consult its own memory for answers to your questions. It can’t do legal research, review cases, review documents, and then answer your questions based on that information. The AI’s output won’t be based on reliable information, and therefore won’t meet the level of accuracy lawyers need.

But imagine an AI as skilled with language as it is today—better, even—that also has access to accurate information, that’s been trained with the right inputs and refinements, and that’s private and secure enough to be trusted with confidential data. That’s when things will get really interesting. 

Reality check: There eventually will exist an AI that can be relied on for professional use by lawyers.