Litigation Nation

Jack Sanker: Welcome Litigation Nation. I'm your host, Jack Sanker, along with my co-host Luke Benke. Today we're talking artificial intelligence and the law. Today's stories of, first, how will artificial intelligence affect law firm business models according to a recent. Law 360 Analysis AI could kill the billable hour model, and firms may be unprepared for what happens next after that.

Jack Sanker: On the flip side, what happens if you try to use chat g p t to write your next brief? A lawyer in New York recently found out the hard way that it may not be a great idea as a result of that. A federal judge in Texas outright banned the use of AI for brief writing in his courtroom. And

Luke Behnke: there's another law school admissions test in

Jack Sanker: Arizona.

Jack Sanker: And as our last bit, I'll leave you with some predictions and analysis of what the immediate future will hold for AI in the law.

Luke Behnke: All that and more coming your way. Here's what you need to know.

Jack Sanker: I'll start with the gloom and doom first. According to a recent analysis by Jonathan Cole, general Counsel at Melody Capital Partners, the billable hour may go the way of the dinosaurs as the use of AI expands in the legal field. Uh, writing in law 360 Cole speculates about how AI might automate some of the.

Jack Sanker: Grinding work that attorneys often bill their clients for by the hour, quote, take for example, document review for discovery or due diligence, which associates and clients both dread but for different reasons. Algorithmic document review obligations turn through many thousands of pages per hour and accurately identify key documents and provisions based on semantic searches, which means that the algorithm can determine the intent and the.

Jack Sanker: Contextual meaning behind a search query, not just key terms or phrases. Indeed, studies have shown that a well-trained algorithm outperforms human reviewers in accurately identifying responsive and non-responsive documents in discovery and at a small fraction of the cost case. Text. Co-counsel suite of GPT four powered applications can, among other things quickly and accurately, diligence documents, review contracts, research, and draft legal me Miranda, and prepare deposition questions.

Jack Sanker: Personally, I think that's probably the easiest use case for AI in the legal profession. We already use boilerplates and to some extent algorithmic research makes total sense that AI would continue to make those processes more efficient. You can argue both sides, but I suspect AI will be incorporated into firm life sooner or later.

Jack Sanker: Now, the piece gets interesting as it moves into the business modeling of law firms automating certain legal work like document review can and probably will reduce the amount of billable hours to go around. Here's an example from the piece. Well, to illustrate the point, let's look at. A not so imaginary invoice, the client is asked to trust that the six hours recorded by junior lawyer each day for three days to quote, attend.

Jack Sanker: Researching and writing memorandum accurately captures the time spent in further, except that at the rate of $540 per hour, the invoice work for a first draft is worth 9,000. $720, the AI legal assistant will be able to research and write a first draft of the memorandum in minutes at a fraction of the cost of a lawyer.

Jack Sanker: Of course, the research will have to be checked in the draft memorandum, reviewed and improved by a lawyer, but there is now a $9,720 hole in the law firm's income statement. How will law firms manage this erosion in revenues unquote. So yeah, that is a bit concerning. I bill by the hour and I don't do much document review these days, but there were definitely times, especially as a lower associate and as a, a young attorney, that many of my invoices were filled with doc review bills.

Jack Sanker: Couple that with trends of reducing billing time for things like travel and in-person appearances with kind of the industry-wide shift to Zoom and uh, remote hearings. And they're just made. Be less billable. Time to go around. Overall, the piece goes on to speculate how staffing and firms may shake out.

Jack Sanker: In this scenario, quote, law firms could rightsize the number of salaried lawyers and partners depending upon the number of attorneys to be laid off. This could be tough choice because unlike businesses that sell products, labor, the billable hours of its attorneys is the product. Law firms certainly called a herd every year by pruning partners, firing the underperformers and letting senior associates know that there will be no ups.

Jack Sanker: So they're out. The ranks are replenished with recruits, fresh outta law school and lateral hires. But now that lawyers who in the past could earn their keep by grinding will have their billable hours reduced at the hands of AI legal assistance. If a law firm adopted AI Technologies does not fire lawyers, will the salaried attorneys be able to meet their annual targets of say, between 2020 500 billable hours?

Jack Sanker: There will be a revenue floor below which the continued deployment of a lawyer is not economic for the partners. Formerly salaried lawyers will join the gig economy as contract attorneys unquote. Yikes. Now as we're gonna see in the next story, AI's current state may not be advanced enough to really hurt lawyers.

Jack Sanker: Uh, I think the jury's still out on that, but I think it's a fair prediction that over the next 10 years, the cost of knowledge in general and the cost of expertise will dramatically fall. And lawyers who operate on the billable hour, I mean, that's our product. That's what we sell. The time that it takes to apply our knowledge and our expertise to a client matter Overall, I think law firms will probably get creative with pricing.

Jack Sanker: Maybe we'll stop using the billable hour and move on to something else, whether it's a flat fee arrangement or something else, I don't know. But for now, if you're bullish on ai, you should also be thinking about what it means for knowledge professionals such as lawyers.

Jack Sanker: Now, on the other hand, maybe we shouldn't worry about AI much at all because as one poor lawyer in New York recently found out, he tried to rely on chat G P T to write a brief for him, and it worked out pretty poorly. I'm not gonna mention the guy's name, but it's easy to find if you want to. Anyways.

Jack Sanker: Personal injury lawsuit filed in the Southern District of New York, uh, against a Columbian airline. The plaintiff's attorney apparently used chat G p T to do some legal research to help respond to a motion to dismiss. Now, according to the affidavit filed by the plaintiff's lawyers, and I'll explain to you in a minute why there's an affidavit quote.

Jack Sanker: It was in consultation with generative artificial intelligence website chat gpt at your affiant, which is the lawyer did locate and cite the following cases in the affirmation in opposition submitted, which this court has found to be non-existent. And then it goes on to list six cases. So this lawyer asked chat G P T to find case law.

Jack Sanker: Dec cite in response to a motion to dismiss and chat. G B T provided six completely fictional made up cases complete with blue book style citations and even fictional holdings and dicta, and this poor lawyer probably says, wow, I've been hearing about how cool and innovative this AI stuff is. It's really an amazing research tool.

Jack Sanker: I'm so grateful that I was able to find these super useful cases so quickly, and they're completely fake. I mean, come on. The judge is considering sanctions against his attorney. In response, they filed the affidavit that I just mentioned. Not only that, but get this in an attachment to the affidavit. The lawyer provides a printout of another conversation with chat G P T wherein the lawyer asked chat, G P T, whether the cases that were.

Jack Sanker: Fake, were bogus or not, and chat. G P t reassures the lawyer over and over again that the cases are real and can be found in reputable legal databases such as LexisNexis and Westlaw unquote. So chat b d doubled down on these fake cases even when confronted. I don't know, man. I actually feel kind of bad for the lawyer who utilized this.

Jack Sanker: I mean, you obviously have to read the case that you cite for a reason. But in any event, maybe we have a little while before we need to be worrying about AI stealing our jobs.

Jack Sanker: All right. So if I, I'm not sure if this was directly in response to the New York fiasco, but the Texas judge in the Northern District of Texas recently amended his standing order to require attorneys. That are appearing in his courtroom to certify that they did not use AI to draft any of their filings.

Jack Sanker: And this order also brings up some other really interesting points that practitioners, you know, should be thinking about. I'll quote from the standing Order Now quote. These platforms are incredibly powerful and have many uses in the law form divorces. Discovery requests suggest errors and documents, anticipated questions and oral argument.

Jack Sanker: But legal briefing is not one of them. Here's why. These platforms in their current states are prone to hallucinations and bias on hallucinations. They make stuff up, even quotes, and citations as an aside, that I think is clearly referring to the New York case. Going back to the order. Another issue is reliability or bias.

Jack Sanker: While attorneys swear an oath to set aside their personal prejudices, bias, and beliefs to faithfully uphold the law and represent their clients, generative artificial intelligence is a product of programming devised by humans who did not have to swear such an oath. As such, these systems hold no allegiance to any client.

Jack Sanker: The rule of law or the laws and constitution of the United States or is addressed above the truth unbound by any sense of duty, honor, or justice. Such programs act according to computer code rather than conviction based on programming rather than principle. Any party believing a platform has the requisite accuracy and reliability for legal briefing may move for leave and explain why.

Jack Sanker: I think that's really well put. The last part about how artificial intelligence have no allegiances to the client is just really interesting. Yeah. If an AI is actually drafting something, I mean, isn't that technically the practice of law? We covered some of the issues with this way back in, uh, episode 35 when we talked about do not pay, which was, uh, quote unquote, Artificial intelligence robot lawyer, um, where we detailed those issues with AI lawyers practicing law.

Jack Sanker: I wouldn't be surprised if we saw more judges try and get ahead on this with their own standing orders, so,

Luke Behnke: According to Reuters Legal, a law school entrance exam developed by the University of Arizona, James E. Rogers College of Law has one approval from the American Bar Association, which means it can be used along with or in place of the traditional law school admissions test, LSAT as it's commonly known.

Luke Behnke: And the G R E for law school admissions, there is a catch. The program called JD Next currently applies only to Arizona law. Applicants other schools evidently can request it, but Arizona Law Dean Mark Miller told Reuters that no other schools have taken that step yet. Interestingly, the University of Arizona in 2016, Was the first law school to start using the G r E in law school admissions, along with the lsat.

Luke Behnke: Now, the purpose of JD Next, which includes an eight week online course for prospective law students, is to give participants a taste of law school and gauge their ability to learn that material. It also, uh, purportedly aims to capture participants' law school aptitude without reproducing the racial score disparities.

Luke Behnke: Seen on other standardized tests. That's according to Dean Miller. The 2019 study of the LSAT found the average score for black test takers was 1 42, which is out of a possible 180, uh, compared with 1 53 for white and Asian test takers. Now reducing score disparities will likely if it hasn't already become a priority for law schools.

Luke Behnke: Should the US Supreme Court rule this month that affirmative action and college admissions is unconstitutional for its part. The Law School Admission Council, which produces the lsat, said in a statement that its exam is empirically proven to be the single best predictor of law school's success. And is a powerful tool for diversity.

Luke Behnke: It also added that it's 2022 cohort of first year law students was the most racially diverse on record.

Jack Sanker: All right, here's another angle on this that I've been thinking of. I kinda went down my own rabbit hole on another technological leap that revolutionized how people advertise, which was the advent and widespread adoption of robocalls in the 1980s. A strict history of like what actually happened in the eighties is hard to find, but from what I can tell, the original practice of telemarketing changed in the eighties when callers could play prerecorded messages to folks and numbers could be auto dialed by computer programs to allow for mass call campaigns.

Jack Sanker: That became easier obviously once computers became more widespread and affordable, which, you know, this was the eighties. It got so bad that in 1991, the government took action and the Telephone Consumer Protection Act set a number of laws in an attempt to protect people from robocall made it illegal for companies or individuals to make prerecorded robocall with the intent of trying to sell you something.

Jack Sanker: For example. However, automated messages by themselves are still legal and they're still quite prevalent. In 2019, it was estimated that about 50% of all calls placed were spam calls. So why don't I bringing up robocall? Well, I think there is really a lot of the practice of law that kind of boils down to letter writing, putting people on notice of things.

Jack Sanker: Things like demand letters, cease and desist letters. Generally, you need a lawyer for those things, but imagine mass campaigns of cease and desist letters or demand letters. Anyone could file a complaint anywhere for any time, for any reason. It might be frivolous. It's gonna get dismissed immediately, but there's no real roadblocks to filing one.

Jack Sanker: So I am somewhat worried about that. The biggest change we might see will be an uptick in spammy junk litigation. Things like slap suits or just frivolous lawsuits meant to intimidate and frustrates someone you don't like. Dumping mountains of discovery responses or request to admit or subpoenas on an opponent, which by the way is already a legitimate and somewhat common tactic in certain cases now, just got a whole lot cheaper and faster and easier to do.

Jack Sanker: Like it or not, I think that the prohibitive costs of legal services is probably the single biggest gatekeeping mechanism that we have against frivolous filings and abuse litigation practices. I. And if you take that away as generative AI seems like it could do. I do see a lot of potential for abuse, not robocalls, but robo cases, if you will.

Jack Sanker: All right, everyone. That's the show. As you know, you can find us on Apple Podcasts, Spotify, YouTube, wherever you get your podcasts. We publish every other Tuesday, and other than that, we'll talk to you in two weeks.