Join Earl Duby in this episode of Big Reports in 5 Minutes as he unpacks the key insights from the New York State Bar Association’s comprehensive 85-page report on artificial intelligence in the legal sector.
Earl simplifies the extensive document to bring you the most crucial points, focusing on AI’s potential to revolutionize legal practices and address the ‘justice gap’ for underrepresented communities.
Discover the benefits and risks of AI for legal professionals, the ethical implications, and the report’s recommendations for integrating AI responsibly into legal frameworks. Whether you’re a legal practitioner or simply interested in the intersection of technology and law, this summary provides valuable insights without the need to tackle the full report.
Link to full report: https://nysba.org/app/uploads/2022/03/2024-April-Report-and-Recommendations-of-the-Task-Force-on-Artificial-Intelligence.pdf
Transcript:
(0:05) Hello everyone, this is Earl Duby. (0:07) Back again with another episode of Big Reports in 5 Minutes. (0:11) Give or take a few minutes.
(0:13) So this episode, I found this really cool report. (0:17) It’s from the New York State Bar Association. (0:20) So they have this task force on artificial intelligence.
(0:23) And they put out this 85-page report. (0:27) And so you don’t have to read 85 pages. (0:29) I read 85 pages.
(0:31) And I’m going to give you the high points. (0:35) So they first start out like, (0:37) it’s a really good summary of how we got here (0:41) from an artificial intelligence standpoint. (0:43) So there’s some good detail, some good history on computing (0:46) and how that’s evolved into artificial intelligence, (0:50) large data lakes, LLMs, large language models, (0:54) and things like that.
(0:55) So you kind of got to work through the background (0:58) to get to kind of the meat of the report, (1:00) which is they get into how artificial intelligence (1:05) and chatbots and things like that (1:06) are about to revolutionize the legal profession. (1:11) So we think about it in terms of,(1:13) hey, how are students going to use it to write reports? (1:17) And how are we going to put doctors out of business? (1:22) Things like that. (1:22) But the legal profession really is going to get(1:27) quite a bit of revolutionizing (1:30) through this artificial intelligence.
(1:33) So they get into the benefits and the risks (1:37) of how lawyers can use artificial intelligence, (1:42) but they focus on it in a different way to start off with. (1:46) They look at it from the perspective of clients (1:49) because they spend a lot of time talking about (1:53) how there is like a two-tiered legal system today (1:56) between the people who can afford good representation (2:00) and the people who cannot afford good representation. (2:02) And they actually have a very interesting statistic in here, (2:06) which I want to take a note of here.
(2:09) So in the U.S., if you look at all the people (2:12) that live below the poverty line, (2:16) 92% of those people, and there’s about 50 million of them, (2:19) 92% of them can’t get access to legal representation (2:24) just because it’s out of their reach (2:27) from a financial perspective. (2:29) And so this report calls that the justice gap. (2:33) And they believe that artificial intelligence (2:35) is one way to reduce this justice gap (2:39) by making legal representation, (2:42) or at least legal advice, (2:45) more available to people that have lower incomes.
(2:48) But then they turn right around and say that that’s, (2:51) that actually could create a larger justice gap (2:55) because the people that are living below that poverty line (2:59) typically don’t have access to technology (3:02) and may be a little bit more computer illiterate (3:05) than the people that are in the higher income levels (3:09) and have access to that legal representation.(3:12) So that’s what’s kind of cool about this report (3:14) is they show both sides of this argument (3:17) and then you’re going to have to figure out (3:19) where you think you stand on that. (3:22) So if you’re a lawyer, (3:23) and I would really suggest you pick this report up (3:26) and we’ll put the link at the end here (3:28) so that you can go download the report.
(3:31) And even though it’s out of New York, (3:33) I think there’s a lot of things in here (3:35) that apply just generally to the legal profession, (3:39) whether it’s in the U.S. or even globally. (3:42) Then they move into the next section of the report (3:45) is called the legal profession (3:47) and the impact on the legal profession. (3:50) And they talk about two rules of professional conduct,(3:54) so the RPCs.
(3:56) So they talk about RPC 1.1 and RPC 5.3. (4:01) Now 1.1 talks about the rules of professional conduct. (4:06) And then RPC 5.3 talks about the supervisory obligation (4:13) that lawyers have for non-lawyer, (4:18) people that are doing non-lawyer work for legal teams. (4:23) And so that would definitely include (4:25) artificial intelligence and chat box.
(4:28) And this part of the report really struck me (4:32) to the point where I wrote a blog about it.(4:35) So if you follow this link, you can get to the blog (4:37) where I go into this in more detail.(4:40) But this section of the report talks about (4:45) even if you are offering up legal advice,(4:48) so think about like these Internet sites (4:51) where you can go and get a will built out (4:54) or you can get a trust or some sort of, (4:58) you can start a business with an online service.
(5:02) Well, this is talking about (5:03) what the lawyer obligation is for that advice. (5:08) And the fact is, you know, (5:10) in terms of the American Bar Association, (5:12) you do have to oversee that. (5:13) So even if there is technology involved, (5:17) you have to have legal oversight (5:19) over what that technology is doing.
(5:21) And the main reason for that is because (5:24) in many cases right now, (5:27) these LLMs and chat box are giving out wrong advice. (5:32) You know, there are tons of hallucinations in there. (5:34) They actually cite a statistic in here (5:39) where the chat box were inaccurate (5:43) a majority of the time when answering a legal question.
(5:47) So over 50% of the time, (5:49) they’re wrong just answering a general legal question. (5:51) And they’re wrong 75% of the time (5:53) when answering a question about a specific court case(5:56) and how that court ruled on that court case. (5:59) So obviously, there’s a lot of variability(6:03) in the answers and whether or not you get a right (6:06) or a wrong answer.
(6:07) And so the Bar Association is saying (6:10) you need to have some legal oversight of that(6:14) just to make sure that your clients (6:16) aren’t getting wrong advice and being, you know,(6:19) led down a wrong path because of that. (6:21) And then they bring up the next section (6:23) talks about judicial oversight (6:26) and judicial responsibility. (6:28) So what do judges need to get involved in? (6:30) You know, they brought up the Mata versus Avianca LLC (6:35) or Avianca Inc.
(6:38) where, you know, some lawyers in that case (6:40) had presented some briefs (6:42) that included incorrect citations, (6:47) court cases that didn’t even exist. (6:49) They were using quotes that never existed. (6:51) And in that case, the judge, you know, (6:55) got pretty upset about that, obviously, (6:57) and made some different rulings.
(7:01) And out of that came the fact that the lawyers (7:04) that had introduced this false evidence(7:06) were fined $5,000. (7:08) So they’re starting to become even a monetary impact (7:11) to the use of AI if it is not done, (7:15) you know, in a responsible way. (7:17) And then the final section of the report (7:19) is just recommendations.
(7:21) There’s some fairly good recommendations in here. (7:24) I think it leaves a lot of room (7:26) for maybe some more specific recommendations. (7:30) But what they do talk about is, you know,(7:33) the legal profession needs to adopt AI guidelines, (7:36) and they actually include a set of proposed guidelines (7:40) in this report.
(7:42) They also talk about focusing more on education (7:45) and making judges, lawyers, law students, regulators (7:49) more aware of just the pitfalls of using AI (7:53) and where it’s useful(7:56) and where maybe it’s not as useful yet. (7:59) So a lot around education. (8:01) And then, you know, along with proposing these guidelines, (8:06) they do say that we should really try to fit(8:08) what we’re doing within the existing legal framework (8:12) and try to use existing laws as much as possible (8:15) to govern AI.
(8:17) And then their final recommendation is just that, (8:20) you know, even if you’re going to use(8:22) the existing set of laws, (8:23) you really should do a gap analysis (8:25) to find out where existing laws aren’t effective enough (8:31) and then figure out how you can fill those gaps in (8:33) with additional laws. (8:35) But they really stress using the existing laws first. (8:39) And then, like I said, they include a set (8:42) of proposed guidelines (8:43) that they would like to see implemented.
(8:45) So this is just for New York, (8:46) but, you know, I’m sure the American Bar Association(8:49) could take a look at that (8:50) and see if maybe it applies more nationally. (8:53) And then they include a pretty good list (8:56) and a nice detailed discussion (8:59) around all of the legislation that they looked at (9:01) as they were preparing this report. (9:03) So, you know, a lot of the things (9:05) that they say in the report, (9:06) you can go right to the court cases (9:08) and find the support for that, you know, (9:12) and just understand where the people (9:14) who put this report together (9:15) were kind of forming their opinions from.
(9:18) So overall, I think it was a really good report. (9:22) You know, I’m glad I read it. (9:23) Gave me a little bit more context (9:24) around, you know, how we can support (9:26) our law firm clients better (9:29) and then just where the legal profession is headed (9:32) in terms of usage of artificial intelligence.
(9:35) So like I said, if you’re a lawyer, (9:37) definitely pick this report up. (9:38) Take the time to read it. (9:40) I think there’s some really good stuff in here.
(9:42) So until next time, (9:44) read the report, read my blog, (9:47) and stay safe out there. (9:50) Thanks.
