AI In the Legal Profession

Which Way AI Lawyer?

This post brings together multiple other perspectives on the future of AI in the Legal Profession. Predictions, as we all know, are bound to be mostly wrong. However, this post is not about trying to “skate to where the puck is going to be” (Courtesy: Hockey GOAT, Wayne Gretzky). This post is using the predictions offered as an insight into what others also knowledgeable about the law and AI are thinking today.

The Efficiency Guru

First things first, is AI set to become the ultimate efficiency guru for lawyers? As with any prediction, the reality is likely to be something much more mundane than a volcanic change to how we do what we do in the law. The promise of AI for emerging legal tech companies is a world where legal research is as easy as a Google search and drafting a complex motion takes an afternoon instead of days. However, the reality of LLM hallucination, the still murky transparency of various “frontier models” and reliance for lawyers on such tools will remain limited. A frontier model in this context is one that is a general LLM that is produced by one of the well known tech companies, OpenAI, Google, Meta (Facebook) and lately up and comers like Anthropic.

• We're already seeing platforms like ROSS Intelligence use natural language processing to make legal research faster and more intuitive. No more Boolean searches or keyword juggling – you can ask a question in plain English and get relevant results. The reality is this particular AI powered legal research tool is going to quickly become a race to the bottom price-wise. There is nothing unique about the ROSS offering. It can and is easily replicated with hundreds of existing GitHub repositories right now. We shall see how that niche evolves.

• AI can also automate the often-tedious process of document drafting. Imagine feeding your key arguments and relevant case law into an AI system and getting a well-organized first draft of a motion in seconds. This is the most likely first iteration of AI powered legal tools — those that augment, but don’t yet replace lawyers.

• I am currently assisting a skilled domestic relations lawyer in a matter involving claims of financial misconduct. The analysis of mountains of transactions from multiple inter-related accounts is a task that AI is primed for.

All of this efficiency translates to lower costs for clients, increased profit margins for law firms (maybe?), and the potential to serve a wider range of clients. Sounds like a win-win, right?

Breaking Down Barriers to Justice

AI has the potential to make legal services more affordable and accessible to individuals and small businesses. So many claims are left untested because the dollar amounts involved are too small to justify hiring a lawyer or even, sometimes, going through the time and small-ish expense of a small claims matter. While this may be great for clients, it creates a future that lawyers will have to grapple with. Will hourly rates drop when AI can do so many legal tasks that business clients now seek the advice of human lawyers for initially? Will AI lawyer products be faced with litigation for practicing law without a license?

• Think about legal self-help apps like DoNotPay, which guide users through legal processes like filing claims in small claims court, fighting parking tickets, or even applying for visas – all without needing to hire a lawyer. The proverbial camel’s nose is already under the tent here. These tools are going to become more sophisticated and more capable of doing things that lawyers are now being paid to do. What then?

• Every state’s professional responsibility rules and pro bono programs seek, ostensibly, to provide wider access to quality legal services to those who cannot afford them typically. What if an AI tool makes 50% of what lawyers current charge for accessible to everyone at a tiny fraction of those costs? Will lawyers embrace that wider access to legal help or see the encroachment as competition?

This democratizing potential of AI raises important questions about the unauthorized practice of law. Where do we draw the line between helpful guidance and legal advice that requires a licensed attorney? It's a complex issue that courts and regulators will need to address as AI becomes more sophisticated.

Reshaping the Legal Skillset

So, are we all going to become legal programmers in the future? Not quite. While coding skills might come in handy, the real challenge will be learning how to effectively use AI in our legal practice. I can code AI applications and remain a licensed attorney. While programming skills have been helpful in some complex financial or document management tasks, overall, a good lawyer still does not need to know how to code.

• Lawyers need to understand AI's strengths and limitations, becoming skilled at constructing effective prompts and critically evaluating the output generated by these systems. Just like having a new car, to effectively use it you don’t need to know how to build or even repair the engine. But, you do need to know how to use the wipers, defroster, etc. to operate it safely and at its peak of usefulness. The same with AI.

• Law firms will need to invest in training programs to keep their attorneys ahead of the curve, and law schools have a crucial role to play in preparing future lawyers for an AI-driven legal landscape.

• This means integrating AI education into legal research and writing courses, teaching students how to use these tools responsibly and ethically. Authors have been struggling with these issues for years. To what extent does using AI or relying on it convert a human writer into merely a prompter? Same question as to lawyers.

The Ultimate Legal Assistant

No matter how advanced AI becomes, it's not going to replace lawyers entirely. Just like I do not believe software developers will be replaced entirely. My sister, a long time corporate learning and business strategy consultant mentioned the possibility of reverse auctions in the future. Meaning, because of the ubiquity of AI, developers may enter auctions which see the race to the bottom in response to posted jobs by corporations. Who is willing to accept the lowest hourly rate for this task? Sort of like a gig economy expanded to many more jobs than it is today. Will this be the fate of lawyers in the future as well?

The human element of lawyering – critical thinking, empathy, strategic decision-making, the art of persuasion – these seem like irreplaceable qualities. But, are they? Do most lawyers sit down and think carefully about strategic choices or are they specialists who know the law so well, they rarely have to research issues before advising clients? For so many lawyers, they rely on templates, past drafts of motions/briefs to write the current one for the current client. This mimicking of past work is precisely what AI is great at. It’s faster, never tires and has accuracy exceeding humans in that kind of mimicry.

For now AI will be most effective when it augments and enhances human skills, not when it tries to replace them. Some see AI as a force multiplier for productivity. But, is the fear of being replaced by AI just fear mongering? I don’t think so. Lawyers will never just go away, but where litigation and preparation for trial might have taken 10-15 lawyers for a complex case by a large law firm five years ago - today it might take three lawyers leveraging AI to accomplish the same tasks with the same efficiencies and success rates. This means that clients who formerly ignored small firms and solo practitioners may well consider sending them work recognizing they can use AI to achieve what firms deploying 6-12 lawyers can also achieve.

So, what does the future hold for LegalAI?

Predictions are so often wrong, they have become unreliable. But, what are lawyers today concerned with or believing is detectable by predictions. For too many they regard what they do as so specialized and significant, that AI cannot really have an impact on their financial future. For me, that view is naive. There is no industry that will not be touched by AI. Hundreds of aggressive developers (and more every day) are working day and night to invent tools that have the effect of augmenting on the way to replacing humans in many knowledge domains. The law is no different. Yes, lawyers, some of us, use creativity to solve our client’s problems. But, the reality is a much larger percentage of us are more akin to working a sophisticated assembly line.

Consider this very common exchange when two lawyers meet for the first time over coffee or during a break at a CLE.

Lawyer 1: “My name is _________, I work in bankruptcy. What do you do?”

Lawyer 2: “I’m [insert name]. I do divorce law, some criminal law as well.”

This is repeated all over the country every day. Lawyers specialize. We focus on an area to the exclusion of all others. Never handled a DUI? Then, you are nowhere near as good at defending those as someone who handles 50 of them a year. That’s a fact.

Specialization means that much of our work is rote repetition of what we have done for similarly situated clients in the past - especially if those approaches have been successful for past clients. AI can replicate this - easily. That’s the reality today. And, as I have said in other posts. AI and its effect on lawyers is the worst, most unskilled and most clunky now than it will ever be. It will never be worse than it is now and it will get better, not linearly, but exponentially. This is not a lawyer who will start fresh out of law school and learn slowly over years the most effective approaches for clients facing this small set of issues within your specialty. This is a lawyer who has consumed and memorized all the law it was ever taught in its training process. And, it will be able to read, consume, synthesize, remember and improve in weeks, not years. No human lawyer can compete with that aspect of AI as it is already today.

The future in my opinion is not all doom and gloom. If you want to just do what you always have done and not learn more and adapt, then, yes, the future is doom and gloom. If you are careful to stay updated on the latest in how AI can augment what you already do, super power your existing skill set, then, the future is bright and unpredictable - in a good way.