Latest / AI Tools for Practicing Lawyers / Episode 031 AI and Access to Justice

Transcript

Episode 031 AI and Access to Justice · AI Tools for Practicing Lawyers

• 0:03
  SPEAKER_01: Welcome to AI Tools for Practicing Lawyers.
• 0:07
  SPEAKER_01: Practical, no-nonsense guidance to help attorneys put AI power to work in their practice right now.
• 0:14
  SPEAKER_03: What if the biggest promise of artificial intelligence in law isn't making lawyers more efficient?
• 0:22
  SPEAKER_03: What if it's making legal help available to people who don't have a lawyer at all?
• 0:27
  SPEAKER_03: My guest today is Rena Shaw, Executive Director of the Maryland Access to Justice Commission.
• 0:34
  SPEAKER_03: And she put the problem in pretty stark terms.
• 0:37
  SPEAKER_03: There are about 40,000 lawyers in Maryland, but only a tiny fraction of them work in civil legal aid.
• 0:45
  SPEAKER_03: And no matter how many lawyers we add, Rena says there's still an upper limit to the number of people those lawyers can serve.
• 0:55
  SPEAKER_03: Technology may change that equation.
• 0:58
  SPEAKER_03: In this conversation, Rena, Heather, and I talk about AI tools being built inside Maryland legal aid, AI-generated pleadings already overwhelming courts and legal aid lawyers, and whether we need new rules governing who or what can provide legal advice and the possibility of building safe AI systems that can serve people directly.
• 1:25
  SPEAKER_03: Perhaps because the access to justice question isn't simply how can AI help lawyers, it's how many more people can lawyers help because of AI.
• 1:36
  SPEAKER_03: Welcome to another episode of AI Tools for Practicing Lawyers.
• 1:41
  SPEAKER_03: I'm Ron Drescher.
• 1:42
  SPEAKER_03: And I'm Heather Gardner.
• 1:44
  SPEAKER_03: And our guest today is Rena Shaw.
• 1:47
  SPEAKER_03: Rena has spent her career working on the justice gap, and Laley has put work right in the middle of the conversation about AI and legal technology, including Maryland's recent AI and Access to Justice Summit and a major effort examining new tiers of legal services.
• 2:08
  SPEAKER_03: Rena, welcome to AI Tools for Practicing Lawyers.
• 2:12
  SPEAKER_00: Thank you so much for having me.
• 2:14
  SPEAKER_00: It's great to be here.
• 2:16
  SPEAKER_03: So before we talk about AI, could you give us some sense of the access to justice problem that we're actually trying to solve?
• 2:26
  SPEAKER_00: Yeah, happy to do so.
• 2:28
  SPEAKER_00: So our work in the Maryland Access to Justice Commission is within that civil justice space.
• 2:34
  SPEAKER_00: So when you think about the hundreds of TV shows that are out there that generally talk about you have a right to an attorney, anything you can and say will be you used against you in a court of law, you have a right to an attorney if you cannot afford one.
• 2:47
  SPEAKER_00: All of that applies to the criminal justice system.
• 2:50
  SPEAKER_00: And many people don't know that in the civil justice system, which covers things like landlord tenant and housing issues, foreclosure issues, when you're applying for public benefits, or if you're applying for or if you're in an immigration proceeding or a bankruptcy proceeding, all of those are civil justice issues.
• 3:07
  SPEAKER_00: And there's no role to an attorney if you cannot afford one.
• 3:11
  SPEAKER_00: And unfortunately, there are a lot of people in Maryland who cannot afford an attorney at market rate.
• 3:18
  SPEAKER_00: So our studies and our research has shown that about 40% of Maryland households cannot afford basic human needs.
• 3:27
  SPEAKER_00: So they can't afford both to pay for food as well as their rent on a regular basis.
• 3:33
  SPEAKER_00: So 10% of those fall below the federal poverty line, but an additional 30% are sort of the working poor who have jobs and are working, but still cannot afford to meet their basic human needs.
• 3:46
  SPEAKER_00: And those people, by and large, do get involved in the civil justice system.
• 3:51
  SPEAKER_00: So there's national data that shows that about 75% of all low-income people have at least one civil justice problem every year.
• 4:01
  SPEAKER_00: And about 92% of those civil justice problems receive no help or very limited help.
• 4:08
  SPEAKER_00: So what that means is that our courts and other administrative proceedings, there are a lot of Marylanders who are really just going at it on their own without any legal help, without any legal representation.
• 4:21
  SPEAKER_00: And as you can imagine, that means that the consequences are pretty high for them.
• 4:26
  SPEAKER_00: If their stakes, what's at stake for them is their housing or they have debt or they're facing a deportation, pretty high-stakes things.
• 4:34
  SPEAKER_00: It could be child custody, could be public benefits.
• 4:38
  SPEAKER_00: And when they don't have sort of legal help or representation, their outcomes are just really not going to be that good.
• 4:44
  SPEAKER_00: And so the commission exists to ensure that we're really looking at the civil justice system as a whole and we're making sure that the people who need help and the level of help that they need that they're able to get it so that they have fairer outcomes.
• 4:57
  SPEAKER_03: Recently, you had more than a hundred people together for Maryland's AI and Access to Justice Summit.
• 5:05
  SPEAKER_03: What were the conversations that led to creating that panel and that event?
• 5:11
  SPEAKER_00: Yeah, thank you.
• 5:13
  SPEAKER_00: So it was a great event, is one of the first in the country for a commission and a state to focus on AI and access to justice.
• 5:22
  SPEAKER_00: So we're very proud that we could be a leader in this space.
• 5:25
  SPEAKER_00: And really what prompted it is really thinking about if you have this massive justice gap where hundreds of thousands of Marylanders are not receiving the help that they need.
• 5:36
  SPEAKER_00: And you have civil legal aid organizations that are chronically underfunded and cannot, they're meeting the civil legal aid organizations are meeting about 20% of the need.
• 5:45
  SPEAKER_00: That the demand that they're getting, the people at their door, they have to turn a lot of people away.
• 5:50
  SPEAKER_00: And so we're looking at how can this new thing, it's not new anymore, it's been around for a few years, but what do we need to do?
• 5:59
  SPEAKER_00: We can't sit on the sidelines and not look at it as a tool to help.
• 6:06
  SPEAKER_00: How can we improve our own systems within civil legal aid organizations?
• 6:11
  SPEAKER_00: How can we use this tool?
• 6:12
  SPEAKER_00: How can we think about use cases for this tool within our own organizations to ensure that one, we know very clearly that self-represented litigants, people who represent themselves and don't have legal assistance through these organizations, are using these tools.
• 6:29
  SPEAKER_00: Might not be the best way, might actually be not as helpful to the places that they're putting in these motions and applications to the courts and stuff, but they are hungry to use them.
• 6:41
  SPEAKER_00: And so we have to really think organizationally, our community who's out there to try to meet this great need, if there's a way to scale the work that we do, if there's a way to improve our efficiency so that we can spend more time serving clients at a higher level and possibly can serve more of them, I think that's it, it's a responsibility that we have in this space to really talk deeply and look deeply at this issue.
• 7:11
  SPEAKER_03: So what I'm hearing, Rena, is that you're approaching incorporating AI into this process in two ways.
• 7:20
  SPEAKER_03: One way is to help litigants actually kind of understand how to better use AI to further their own interests in legal proceedings if they can't afford a lawyer.
• 7:32
  SPEAKER_03: But the other approach is for public access organizations like Maryland State Bar Association, maybe Maryland Lawyer Volunteer Service, Legal Aid, et cetera, to bring AI into their quiver of arrows, as it were, there, so that they are better prepared to fulfill their mission.
• 7:53
  SPEAKER_03: Am I hearing that right?
• 7:54
  SPEAKER_00: Yeah, that's right.
• 7:55
  SPEAKER_00: And the summit actually just focused on the latter.
• 7:58
  SPEAKER_00: The summit really just focused on how can we think about the civil legal aid organizations, but in other contexts, we have also looked at how do we really think about bringing some of these tools direct to consumer.
• 8:11
  SPEAKER_00: But for the summit, um the whole focus was on the civil legal aid organizations in that community, thinking through use cases, talking to leadership about what are the obstacles, what are the opportunities, and what are some examples about how we can just start playing with this and innovating with this and learning together.
• 8:33
  SPEAKER_00: So that was the focus of the summit.
• 8:35
  SPEAKER_03: So who were the panelists on the summit?
• 8:38
  SPEAKER_00: Yeah, we really tried to focus on Maryland folks and peer-to-peer learning.
• 8:44
  SPEAKER_00: We did a survey in advance, the planning committee, we did a survey in advance of the summit to get a baseline of where is our community.
• 8:53
  SPEAKER_00: And in that survey, we asked some basic questions.
• 8:57
  SPEAKER_00: Does your organization have any enterprise-level AI tools?
• 9:01
  SPEAKER_00: Does your organization have a policy related to AI?
• 9:05
  SPEAKER_00: Some other questions.
• 9:07
  SPEAKER_00: And what we learned from that, we had about 187 staff from different civil legal aid organizations answer that survey.
• 9:14
  SPEAKER_00: And what we learned is that a lot of organizations are very much at that basic level.
• 9:21
  SPEAKER_00: Not many had actually made the foray into having a policy.
• 9:26
  SPEAKER_00: Most did not have enterprise level tools.
• 9:28
  SPEAKER_00: And most were not really even, hadn't really taken a stance in terms of whether their staff should use AI tools, shouldn't use AI tools, were okay with using AI tools.
• 9:38
  SPEAKER_00: And so what we knew is that we needed to start at a very basic level to make sure that we could get everyone on the same page.
• 9:47
  SPEAKER_00: At the same time, you should know that we have one organization, especially Maryland Legal Aid, which is considered a national leader in this space.
• 9:55
  SPEAKER_00: And they've done a lot of work with incorporating AI into their practice.
• 10:00
  SPEAKER_00: So, one, for the commission, we're really thinking about equity.
• 10:04
  SPEAKER_00: How do we bring all of these organizations in this into this space and to like a similar starting level?
• 10:10
  SPEAKER_00: And so, in terms your question was about who are the speakers.
• 10:13
  SPEAKER_00: Well, we started off the day with the the leaders of the Access to Justice Commission.
• 10:19
  SPEAKER_00: Um, we had Vicky Schultz from Maryland Legal Aid.
• 10:23
  SPEAKER_00: We had Donald Tobin, who's the former dean of the University of Maryland Law School, who's part of our um the Access to Justice Commission's Data and Legal Technology Committee.
• 10:31
  SPEAKER_00: And uh our keynote was um Chief Justice Bader from the Maryland Supreme Court, who's also actually a national leader in the Supreme Court justices space.
• 10:42
  SPEAKER_00: As we moved through the day, we had sort of different uh people that were part of our planning committee.
• 10:48
  SPEAKER_00: We had Lindsay Bramble, we had Heather Gardner actually as one of our speakers, and uh we had Donald Tobin.
• 10:55
  SPEAKER_00: Um, and so these are all peers.
• 10:57
  SPEAKER_00: These are people who are in our community, are doing things.
• 11:01
  SPEAKER_00: In the afternoon, we sort of had use cases, we had exercises so that people could actually work through some of these things.
• 11:08
  SPEAKER_00: We had had asked everyone to bring their laptops and they were like, work through this issue, go on this tool and see if you can actually come out with some product.
• 11:17
  SPEAKER_03: So you you kind of turned it into a more of a workshop than just a passive audience receiving information CLE.
• 11:25
  SPEAKER_02: That was a great session.
• 11:28
  SPEAKER_02: After you did that, you had everyone talk about it.
• 11:30
  SPEAKER_02: We did talk about challenges, and yeah, it was a really cool area of the day.
• 11:36
  SPEAKER_00: And If you're gonna call out, I'll just call out some people that I want to make sure I name.
• 11:40
  SPEAKER_00: So John Jeff Cott from Maryland Legal Aid put all those use cases together.
• 11:44
  SPEAKER_00: Um, and then we also had Angela Tripp from a national organization.
• 11:48
  SPEAKER_00: She was at um the Legal Services Corporation, but now she's at the National Center for State Courts.
• 11:53
  SPEAKER_00: And they all can Angela in the beginning provided sort of this baseline of information.
• 11:57
  SPEAKER_00: What is AI?
• 11:58
  SPEAKER_00: What should you do about it?
• 11:59
  SPEAKER_00: It's a prediction model, not like something that you can rely on.
• 12:03
  SPEAKER_00: It's not a search engine.
• 12:04
  SPEAKER_00: So there were some basic things that we needed to communicate.
• 12:07
  SPEAKER_00: And so we moved from that into let's do some practical, yeah.
• 12:12
  SPEAKER_03: Okay.
• 12:12
  SPEAKER_03: Well, you just opened the door for something, Rena.
• 12:15
  SPEAKER_03: All right.
• 12:16
  SPEAKER_03: You said it's a it's a prediction model, not a search engine.
• 12:22
  SPEAKER_03: How do how do you use that that kind of distinction to communicate the value and the limitations of the tools?
• 12:35
  SPEAKER_00: Yeah, I mean, I think some things that we learned, for example, from our survey in terms of the concerns that people have.
• 12:42
  SPEAKER_00: People have a concern about accuracy.
• 12:44
  SPEAKER_00: That's one of the biggest concerns that people have, especially in the legal community.
• 12:48
  SPEAKER_00: They're very concerned about what you read in the news, hallucinations.
• 12:52
  SPEAKER_00: And lawyers, I think it's more of a problem where lawyers are not doing their due diligence and checking.
• 12:59
  SPEAKER_00: But these are the kinds of news articles that come out where people are concerned that are they, if they go to AI, are they going to get something that is inaccurate or is is hallucinated?
• 13:10
  SPEAKER_00: And I think understanding that whatever you're asking AI to do, and another issue was I'll just say, was the environmental cost of things.
• 13:18
  SPEAKER_00: So understanding, okay, you're asking it to do something and it's breaking it up into these tokens, and then it's coming out with some prediction.
• 13:27
  SPEAKER_00: So like every time you do a search, you might not get the same thing, or every time you ask it to do something, you might not get the same thing.
• 13:35
  SPEAKER_00: But that even if you don't have that, so you get over that hesitation, explaining like closed systems, how they can reduce sort of error or inaccuracies, and being able to say that yes, these are all, this is the way it works.
• 13:52
  SPEAKER_00: So that's why you'll see these differences, but there is still a way to make it safe and responsible for our purposes and add in protection.
• 14:00
  SPEAKER_02: Every speaker throughout the day, Ron, discussed the limitations and the need for human verification and still having human judgment in the loop and where that belongs, etc.
• 14:10
  SPEAKER_02: So every person that got up and spoke during the day, I think reiterated that multiple times.
• 14:16
  SPEAKER_03: So, you know, to me, as far as if you're going to implement AI in a legal setting, a legal aid service or an access to justice organization is similar to a law firm.
• 14:28
  SPEAKER_03: And they have to, they have to make a lot of the same kind of decisions.
• 14:33
  SPEAKER_03: So if you're trying to establish a baseline use of AI in the constituency that you are looking at, what are the first three things that these organizations need to consider?
• 14:46
  SPEAKER_03: Because people get paralyzed.
• 14:48
  SPEAKER_03: Well, I don't know, I got this free tool, so so I could afford it, but is it the right tool and am I using it in the right way?
• 14:57
  SPEAKER_00: Yeah, I mean, I think a very clear need that came out was that people really want to work on their policy first.
• 15:03
  SPEAKER_00: We had sort of a lunch session where we talked to some leaders and understood how they are approaching this within their organization.
• 15:12
  SPEAKER_00: And it seemed like Vicki Schultz from Maryland Legal Aid, who's long in this process, what she said was you really have to come together with your organization and figure out what your values are and make sure that in your AI policy, you're reflecting those values and you're saying very clearly, like, these are our boundaries, this is what we what we need in every contract or MOU or whatever you do with an organization, or if you're getting an enterprise version of a tool, these are the safeguards that we need.
• 15:42
  SPEAKER_00: So I think the leadership was very clear that they want to be able to have a clear policy before they go in and start doing anything.
• 15:51
  SPEAKER_00: The other thing was that they really want to learn more about the tools.
• 15:54
  SPEAKER_00: There was not enough awareness or knowledge about what are trusted tools, which are vetted tools, what tools are good for what, and and things like that, some basic information, because maybe not all tools are good for all things, and maybe not all things require the same level of protection.
• 16:13
  SPEAKER_00: So if you're working on a legal matter for a client, which a lot of it we didn't really talk about that, we were thinking if you want to do some development work, if you want to do research on foundations, because all of these organizations need to fundraise, if you want to have some HR functions, if you want to do communications, that's not gonna require the same amount of protection.
• 16:35
  SPEAKER_00: And you can think differently about that than you can, for example, maybe the development case, if you're reporting, that could be different.
• 16:42
  SPEAKER_00: But if you're actually thinking about using a tool to help a client and you have to protect the client information.
• 16:49
  SPEAKER_00: So I think there's like the differences, but I think another need that was clear is people really want to understand the tools, what it's good for, what level of protections they offer, and how they can feel sort of safe in using some of these tools.
• 17:05
  SPEAKER_03: You know, in May, Anthropic came out with Claude for Legal, and it made a pretty big splash in our industry.
• 17:14
  SPEAKER_03: And for the most part, it's a suite of tools that are designed for big law.
• 17:21
  SPEAKER_03: But Anthropic did make a nod towards access to justice by including a couple of connectors and a couple of skills within the suite that they were offering for pro bono legal aid type organizations.
• 17:40
  SPEAKER_03: All right.
• 17:41
  SPEAKER_03: Have you looked at those?
• 17:42
  SPEAKER_03: Did you talk about those at all in your in your panel?
• 17:46
  SPEAKER_03: Have you looked at them and and and analyzed them?
• 17:50
  SPEAKER_00: Yeah, that is a great question.
• 17:51
  SPEAKER_00: And I uh unfortunately have not.
• 17:54
  SPEAKER_00: I I did encounter Claude for Legal when we were working.
• 17:58
  SPEAKER_00: So Heather and I actually are working on this task force together.
• 18:02
  SPEAKER_00: It's called the Innovations and Tiered Legal Services Task Force that the commission has been co-leading with the Maryland Judiciary and the Maryland State Bar for the past, I would say, 18 to 20 months.
• 18:14
  SPEAKER_00: And in that, we have a committee that looked at legal technology.
• 18:17
  SPEAKER_00: And that committee, instead of looking at how we can put legal technology and AI in the civil legal aid organizations, we were actually looking at could these legal tech tools serve consumers directly?
• 18:30
  SPEAKER_00: And we did have several presentations through that task force work on cloud for claud for legal and saw the capability.
• 18:39
  SPEAKER_00: But to be honest, I am not aware of these connectors and skills for pro bono and legal aid.
• 18:47
  SPEAKER_00: Maybe I made the assumption because I know that it was presented to me as like a big law tool.
• 18:53
  SPEAKER_03: Yes.
• 18:54
  SPEAKER_00: So I made the assumption that maybe this is not really for us, but a conversation that we've had in our community is that some of these tools we do need at a different cost structure than is made for big law, which is a huge, we don't want to increase the justice gap and not have access to these tools that other law firms are able to use and big law is able to use.
• 19:15
  SPEAKER_03: Well, you know what?
• 19:16
  SPEAKER_03: That kind of leads me to something that I was thinking about.
• 19:19
  SPEAKER_03: If anthropic is going to make a nod towards access to justice by releasing de minimis tools as part of a larger launch, does anthropic, open AI, Google, Grok, do any of them have programs to make their tools more easily affordable or accessible to organizations like yours?
• 19:45
  SPEAKER_03: And have have you talked to them?
• 19:46
  SPEAKER_03: Have you tried to convince them that they should?
• 19:50
  SPEAKER_00: So we're part, we went, for example, we went to a conference that Stanford University put on last year.
• 19:57
  SPEAKER_00: It was on AI and access to justice, but it was a national summit and it brought together people from all the states.
• 20:03
  SPEAKER_00: And there is so it's like those kinds of national organizations, the Legal Services Corporation and some national organizations who work in this space that are saying that they want to, they are attempting to communicate with these larger sort of LLM organizations to say it shouldn't be one, it shouldn't be a state-by-state solution.
• 20:23
  SPEAKER_00: Like this, our issues are sort of nationwide.
• 20:27
  SPEAKER_00: Every state has an access to justice crisis and a problem.
• 20:30
  SPEAKER_00: And we really do think that these tools should be made available and affordable and even free.
• 20:36
  SPEAKER_00: I mean, honestly, we're a small percentage.
• 20:41
  SPEAKER_00: They could be given to us in a pro bono basis, but we have talked a lot about how do we engage.
• 20:47
  SPEAKER_00: We haven't done it directly.
• 20:48
  SPEAKER_00: We are uh relying sort of on some of these national leaders to bring about something that would serve all of our different commissions across the country.
• 20:59
  SPEAKER_03: How are the national leaders engaging the open AIs, anthropics, et cetera, of the world?
• 21:06
  SPEAKER_00: I mean, that's a good question, Ron.
• 21:09
  SPEAKER_00: That's a good question.
• 21:10
  SPEAKER_00: I can't say that I have a good answer because clearly we don't have a solution yet.
• 21:16
  SPEAKER_00: So I know efforts are being made, but I have to say that that that piece is not directly something that I have followed deeply, but I do have a conference that I'm going to next week.
• 21:26
  SPEAKER_00: And that is one of the questions that I plan to ask.
• 21:30
  SPEAKER_02: And I want to say uh that I think that the earlier discussion kind of is the reason for that, right?
• 21:36
  SPEAKER_02: When we get when you got in that room with 187 people from different Maryland, you know, civil organizations, what we saw is they're at a place where they're ready to put together a policy, Ron.
• 21:49
  SPEAKER_02: So we aren't at the stage of conversation you're at yet.
• 21:53
  SPEAKER_02: Because, but I I do want to also say that in the legal field and what we discuss on this show week after week is the cost.
• 22:00
  SPEAKER_02: And the hesitance that attorneys have, and even rightfully so.
• 22:04
  SPEAKER_02: Access to justice is one area in the Maryland Legal Forum that is going full speed ahead, still doing it cautiously, but they're not in the camp of we're not going to use these tools, right?
• 22:16
  SPEAKER_02: Like we hear so often.
• 22:18
  SPEAKER_02: So they are ahead of the game, even though they're not quite worth, you know, you're asking right now.
• 22:24
  SPEAKER_03: I'm curious about the efforts to educate the public on how to use AI tools on a pro se basis, as opposed to having having a lawyer from an organization such as the MSBA or legal aid.
• 22:43
  SPEAKER_03: What does that outreach look like?
• 22:45
  SPEAKER_00: That's a great question.
• 22:47
  SPEAKER_00: It that is not something that we have, we know it's a problem.
• 22:53
  SPEAKER_00: We know it's a problem because self-represented litigants are putting a lot of their personal information into open systems and asking for legal advice.
• 23:04
  SPEAKER_00: What we know is that people are so hungry and so desperate for answers.
• 23:11
  SPEAKER_00: They do not want to go to a website and get some general answer that doesn't specifically address their question or concern.
• 23:22
  SPEAKER_00: What they really need is an answer.
• 23:24
  SPEAKER_00: And this just points large to the larger access to justice crisis is that we don't have enough people, we don't have enough organizations, we don't have enough supply to provide people with what they need.
• 23:37
  SPEAKER_00: So they're going to where they think they can receive some answer.
• 23:42
  SPEAKER_00: The challenge there is that, of course, there's security concerns, they're sharing their information from a court's perspective.
• 23:49
  SPEAKER_00: There's also, if something was putting to a large LLM, can that be used in some way?
• 23:56
  SPEAKER_00: So there's all kinds of complications with self-represented litigants doing that.
• 24:01
  SPEAKER_00: But also from the perspective of, let's say, the court help center, which accepts all Marylanders, regardless of income.
• 24:10
  SPEAKER_00: You can go to a court help center and ask for help.
• 24:13
  SPEAKER_00: And what we are learning from them is that they are receiving more and more motions or pleadings that are developed by AI LLM models, and people are walking in with them, and it's pages and pages and pages of things.
• 24:29
  SPEAKER_00: And there's probably kernels of truth in there.
• 24:32
  SPEAKER_00: There's probably some statute and case law that is applicable to their case.
• 24:36
  SPEAKER_00: However, the amount of things that the human has to do to review and then to say, well, some of this is accurate, but a lot of it is not, and it wouldn't apply to your case, is actually adding so much work.
• 24:50
  SPEAKER_00: And honestly, the self-represented litigants are not believing what the human attorneys are telling them.
• 24:55
  SPEAKER_00: They're saying, no, I was given this information and advice from this tool, and so you are wrong.
• 25:02
  SPEAKER_00: And so it's creating this issue.
• 25:04
  SPEAKER_00: The Access to Justice Commission is now, we think one solution that we addressed in our innovations task force is that we cannot sit on the sidelines and not develop tools that can be safe and closed and that can actually help self-represent litigants.
• 25:23
  SPEAKER_00: So that was our whole effort is to say, can we give some authorization to a legal tech tool to be able to give legal advice in a very low-risk area of law and where we can have a lot of predictability?
• 25:39
  SPEAKER_00: And uh because the LLMs are giving legal advice, they're crossing unauthorized practice of law, and there's no sort of repercussion for that.
• 25:49
  SPEAKER_00: And so we're saying, why not give a waiver of UPL to allow us to build safer tools for self-represented litigants that we can actually test and we can actually vet, and there's a process by which we accept them so that they can use them.
• 26:05
  SPEAKER_00: They want something that they can use that's easy to use.
• 26:09
  SPEAKER_00: However, beyond that, there's a lot more needed.
• 26:13
  SPEAKER_00: And I was just speaking with a colleague at the People's Law Library, because the People's Law Library is sort of based in the judiciary.
• 26:21
  SPEAKER_00: They sort of put out all of this information for anyone in Maryland who it's like the Web MD in the health space.
• 26:28
  SPEAKER_00: It's like if you have a legal problem, a civil legal problem, you can go to People's Law and sort of look for credible and quality answers.
• 26:37
  SPEAKER_00: And they are looking a lot at do they need how they can put some guidance up for self-represented litigants?
• 26:45
  SPEAKER_00: How do you make your own websites as accurate as possible and as helpful as possible for AI tools to pull the correct information?
• 26:55
  SPEAKER_00: So different people are looking at these different things, but I think our solution is that we should develop some safe tools for self-represented litigants to use so that they have access and can get those answers that they seek.
• 27:10
  SPEAKER_03: You raised an interesting point before.
• 27:13
  SPEAKER_03: And this whole discussion is really fascinating.
• 27:16
  SPEAKER_03: Uh, you talked about how LLMs are giving legal advice and you don't want to be in a position of giving pro se litigants legal advice because that you have all sorts of unauthorized practice of law problems and also just misinterpretation problems.
• 27:33
  SPEAKER_03: But for me, what I'm thinking of as a role for access to justice uh organizations are regular webinars for the public to teach them how to use these tools safely, as opposed to to provide them the tools.
• 27:52
  SPEAKER_03: Because you know what?
• 27:54
  SPEAKER_03: The tools aren't expensive.
• 27:56
  SPEAKER_03: Yeah.
• 27:57
  SPEAKER_03: But if because look, you could use the free version of uh Gemini, ChatGPT, Claude.
• 28:06
  SPEAKER_03: And if you understand that you need to redact the documents that you're going to feed into it, that is a procedure that should keep you safer than if you don't redact.
• 28:19
  SPEAKER_03: And it just seems to me that there are so many opportunities to educate the public in that way, as opposed, you know, teaching them how to fish instead of giving them the fish, and instead of actually giving them the rod and reel.
• 28:33
  SPEAKER_00: Yeah.
• 28:34
  SPEAKER_03: What are your thoughts about that kind of initiative?
• 28:37
  SPEAKER_00: I think that's a good suggestion.
• 28:39
  SPEAKER_00: I know that with the level of apprehension that we're even facing with attorneys, putting legal tech tools in attorneys' hands or AI tools in attorneys' hands where you sort of have qualified people using it, there's apprehension.
• 28:55
  SPEAKER_00: So this conversation that we've started about guiding self-represented litigants, I think there's probably so much more apprehension to broach that, although it's more necessary because it's happening now.
• 29:08
  SPEAKER_00: They're doing it.
• 29:09
  SPEAKER_00: They're doing it.
• 29:10
  SPEAKER_00: They're absolutely doing it.
• 29:12
  SPEAKER_00: And I think your ideas are valid and good.
• 29:16
  SPEAKER_00: Um, I think I never think that there's either or.
• 29:19
  SPEAKER_00: I think this is necessary to educate self-represented litigants.
• 29:23
  SPEAKER_00: I also think that yes, if you're thinking about some some case, you're, you, you're, you don't know how to write a motion.
• 29:31
  SPEAKER_00: And there could be so many more things, prompting or how do you there's a lot of education that can be done so that self-represented litigants are doing it in a way that is actually helpful and all of that.
• 29:43
  SPEAKER_00: But but they're still pulling from laws from like every state, and there could be pulling from laws that are no longer valid.
• 29:51
  SPEAKER_00: So I do think that creating a closed system, if you're landlord tenant law in Maryland, the most current, and you can just go into that and ask your questions and get something out, that is still going to be so much more helpful to you than just like reaching out into the ether.
• 30:09
  SPEAKER_03: I guess where I'm coming from is look, the three of us, we're in the field, we're reading about uh AI developments every day, we're hearing about the thousands of lawyers who've been sanctioned for allowing hallucinations to get into court records.
• 30:26
  SPEAKER_03: And, you know, pro-Heyndigans aren't paying attention to those developments.
• 30:31
  SPEAKER_03: They may or may not be aware that AI hallucinates.
• 30:37
  SPEAKER_03: They may or may not be aware that there are ways to constrain AI from hallucinating in a legal context.
• 30:45
  SPEAKER_03: And so it just seems to me there's so much value to be had in training potential pro se litigants to understand those procedures and to implement uh hallucination control procedures.
• 31:01
  SPEAKER_03: That just seems like a great opportunity.
• 31:04
  SPEAKER_02: There's no such thing as hallucination control procedures, is one of the problems.
• 31:08
  SPEAKER_02: Educating pro se litigants is absolutely a great idea, but you could never ever suggest to them to go use it in blah, blah, blah way, because all you're doing is committing UPL yourself if you do that.
• 31:22
  SPEAKER_02: Number one.
• 31:23
  SPEAKER_02: Really?
• 31:24
  SPEAKER_02: 100%.
• 31:24
  SPEAKER_02: 100%.
• 31:25
  SPEAKER_03: If if I tell a litigant, I'm sorry, if I tell a pro se litigant that hallucination is a danger, and so they have to review every citation and every case, and they have to make sure that's UPL.
• 31:39
  SPEAKER_02: If you're telling them to go use an open AI tool for any sort of legal work, that to me is U.
• 31:45
  SPEAKER_02: I think that by itself is UPL.
• 31:47
  SPEAKER_02: I would never suggest, I mean, that is sending them to something that we've got 92% of this demographic.
• 31:54
  SPEAKER_02: This is not like 5%, right?
• 31:56
  SPEAKER_02: Where you have a chance of actually getting to a majority of these people.
• 32:00
  SPEAKER_02: 92% of civil rights litigants are un or underrepresented.
• 32:05
  SPEAKER_02: Attorneys who know they shouldn't are still making these mistakes.
• 32:10
  SPEAKER_02: So sending a layperson who is probably already working three jobs and can't afford legal representation and has very little time to even try to get paperwork together for legal aid or MBLS, and then trying to get them to take an AI course and retain it and follow it is a little unrealistic.
• 32:29
  SPEAKER_02: Giving them a tool and saying, here, just use this, which should be quick and easy and free for to be developed and handed out.
• 32:38
  SPEAKER_02: Truthfully, I think that's coming.
• 32:40
  SPEAKER_02: I think I envision our self-help clinics now.
• 32:44
  SPEAKER_02: You can walk in and there is a closed system computer that people can go up to look for what's in the court, right?
• 32:50
  SPEAKER_02: Records, etc.
• 32:52
  SPEAKER_02: Yes.
• 32:52
  SPEAKER_02: Another one that has the AI tools that are for that department being there with a safe, closed tool for that person.
• 33:02
  SPEAKER_02: Instead of saying, I know you might get deported tomorrow, but we want you to go home and get get signed up for this AI course and try this.
• 33:09
  SPEAKER_02: I I think that this is a much, you know, more again, attorneys are still using it in the ways that are mucking up the courts.
• 33:17
  SPEAKER_02: I don't think we should send the the the pro se people to do that.
• 33:20
  SPEAKER_02: They are already doing it, of course.
• 33:22
  SPEAKER_03: That's the point.
• 33:23
  SPEAKER_03: There you go.
• 33:24
  SPEAKER_03: They're already doing it, yeah.
• 33:25
  SPEAKER_03: You just you you just made my point, which is I'm not saying to a person who would otherwise never use AI, oh yeah, just go to ChatGPT and enter these prompts.
• 33:36
  SPEAKER_03: No, that that would be irresponsible.
• 33:39
  SPEAKER_03: But we all assume that there everybody's using it anyway because they've heard of it.
• 33:44
  SPEAKER_03: And if they're going to be using it anyway, why not provide them information to help them use it more responsibly?
• 33:51
  SPEAKER_02: And yes, and I think one of the biggest things that came out of that day that is worth discussing, we know that the courts are being mucked up with pro se filings that are AI generated, right?
• 34:02
  SPEAKER_02: Now all of the Maryland civil justice organizations are also being overrun and buried in AI printed out documents as well, where they used to maybe get a one-page thing, they're getting a 20-page, and it's very obviously AI.
• 34:19
  SPEAKER_02: So instead of being helpful, it's actually causing more time to be used and it's causing less people to be able to be served.
• 34:27
  SPEAKER_02: So offering, you know, a demographic of our population who don't have options, offering them education is great, but offering them a tool that will actually meet their needs now is something that the legal justice system should be able to do literally two years ago.
• 34:47
  SPEAKER_03: See, I agree with you, Heather.
• 34:49
  SPEAKER_03: I don't want to offer them the tool.
• 34:51
  SPEAKER_03: I'd rather offer them the education.
• 34:54
  SPEAKER_03: But this is but what you just said opens up another question for Rena, which is okay, Heather just said that it's becoming more and more common for these, you know, access to justice organizations to receive 20-page submissions from underserved clients.
• 35:16
  SPEAKER_03: How do these organizations address that occurrence?
• 35:20
  SPEAKER_03: They get a 20-page manifesto.
• 35:23
  SPEAKER_03: How do you respond?
• 35:25
  SPEAKER_00: Yeah, that is an interesting problem.
• 35:27
  SPEAKER_00: And it's something that obviously these organizations are facing more and more.
• 35:31
  SPEAKER_00: We start with an organization that has limited capacity, whether it's the civil legal aid organization or the court help center staff.
• 35:39
  SPEAKER_00: It is overwhelming.
• 35:41
  SPEAKER_00: And my thing would be maybe there's an AI solution.
• 35:44
  SPEAKER_00: Back a little bit in this conversation, we were talking about this big AI tools and whether we can have some enterprise rates.
• 35:51
  SPEAKER_00: A lot of positive things are that some of these, so like Maryland Legal Aid has been able to create its own.
• 35:57
  SPEAKER_00: They have this thing called MLA GPT, Maryland Legal Aid GPT.
• 36:01
  SPEAKER_00: And they've been able to create their own tools and have their own agents and have figured out how to, yes, you need these agreements with these large language models and all of that.
• 36:11
  SPEAKER_00: But for our issues, we can develop our own things and come up with solutions.
• 36:16
  SPEAKER_00: So I just wanted to highlight like that innovation, maybe it doesn't happen at that very top level with these large LLMs that are really interested in creating tools for big law, but there is sort of this community of folks who are wanting to create and develop things for their own organizations.
• 36:34
  SPEAKER_00: And I would think that something like this, you don't want to use up man or woman power.
• 36:39
  SPEAKER_00: And maybe you put in something that you have a tool to actually be like within this submission, what is legit and what is not?
• 36:48
  SPEAKER_00: And you you put it into several tools and you ask them to spit out what is okay.
• 36:53
  SPEAKER_03: I see it, I see AI talking to AI all over the place.
• 36:57
  SPEAKER_02: Yeah.
• 36:57
  SPEAKER_02: When a really, really easy good example of this, and one of the biggest places all courts in every state and Maryland are strained is family.
• 37:05
  SPEAKER_02: Let's talk about the family court.
• 37:07
  SPEAKER_02: Every county in Maryland has a family self-help clinic, and that is a grossly unrepresented, you know, sector of legal area of the legal field.
• 37:19
  SPEAKER_02: That self-help clinic in each county could have a tool that even just helps people fill out their forms.
• 37:26
  SPEAKER_02: This is something a non-attorney can't even do, right?
• 37:29
  SPEAKER_02: We can't even help a crow se litigant fill out their forms.
• 37:34
  SPEAKER_02: That is repetitious all the time, you know, procedure that an AI tool could easily walk that person through.
• 37:42
  SPEAKER_02: And should anything come up that's irregular, stop them and direct them to someone or or or how to do that.
• 37:48
  SPEAKER_02: But that by itself eats up probably more manpower in those self-help clinics than you know, it ever should have, and could be eliminated with a very easy-to-use tool that's only got all the directions and every fact that they need for those specific forms that have to be filed in every case.
• 38:08
  SPEAKER_03: So, Heather, are you saying that we're gonna have to relax our UPL rules to accommodate access to justice?
• 38:17
  SPEAKER_03: I would never say that.
• 38:19
  SPEAKER_02: I would never say that.
• 38:20
  SPEAKER_02: But our tiered legal services is already looking at that for the future.
• 38:24
  SPEAKER_02: All the states that have done the limited licensing have had to have changes to their UPL laws.
• 38:29
  SPEAKER_02: If Maryland gets there in the future, whether that's with machines or humans, yes, it will have to be changed, altered.
• 38:36
  SPEAKER_02: Maybe not relaxed, but but let me pose this question.
• 38:40
  SPEAKER_00: Yeah, I mean, because Heather is absolutely so we have this entire task force, the innovations task force we've been working for 18 to 20 months, and we have recommendations to have regulatory reform, and we'll see what happens, whether the court accepts those recommendations or not.
• 38:56
  SPEAKER_00: But the larger question is, and related to legal technology, is that these large companies did not have to ask for permission, did not have to ask for a waiver of UPL.
• 39:07
  SPEAKER_00: They just did it, and now it's before us, and we have to deal with it.
• 39:12
  SPEAKER_00: And the only reason that nothing is done is because they're so huge and it's hard to bring them down and say, well, you're crossing UPL.
• 39:20
  SPEAKER_00: I think there's one litigation, one case out there where a state or an attorney general sued one of these large language models to say you've crossed over UPL.
• 39:30
  SPEAKER_00: And yet we spend months, years thinking about whether we can cross this little line so that all of these people can get some little help.
• 39:40
  SPEAKER_00: Meanwhile, there's no consternation on the part of these money.
• 39:44
  SPEAKER_00: Oh, no, no.
• 39:45
  SPEAKER_00: It's happening now.
• 39:47
  SPEAKER_03: It's happening now.
• 39:48
  SPEAKER_03: This is this has become a famous case that got filed, I think, in March of this year.
• 39:51
  SPEAKER_03: Okay, maybe that's the one.
• 39:55
  SPEAKER_03: And it's open AI.
• 39:57
  SPEAKER_03: I think it's Nippon Insurance Company.
• 39:59
  SPEAKER_03: Some woman had a case and it's settled.
• 40:02
  SPEAKER_03: She was represented by an attorney and she collected the money and she went to open AI.
• 40:08
  SPEAKER_03: She went to ChatGPT and she said, Did I get a good deal out of here?
• 40:12
  SPEAKER_03: And and ChatGPT said, No, you got a terrible deal.
• 40:16
  SPEAKER_03: You should fire your lawyer and you should.
• 40:18
  SPEAKER_03: I don't know if this is the exact tax, but this is what ended up happening.
• 40:22
  SPEAKER_03: And you should move to reopen the case.
• 40:25
  SPEAKER_03: So she did all that.
• 40:26
  SPEAKER_03: She moved to reopen the case, and the insurance company, instead of just saying, Do you want another$5,000 to go away?
• 40:34
  SPEAKER_03: fought her tooth and nail every step of the way, ran up$300,000 in legal fees, and now is suing open AI for the$300,000.
• 40:46
  SPEAKER_03: Million in damages, yes.
• 40:47
  SPEAKER_03: Right.
• 40:48
  SPEAKER_03: $10 million punitive damages.
• 40:49
  SPEAKER_03: And let's just be the real government.
• 40:51
  SPEAKER_02: Yeah, but how how that's just it's just ridiculous.
• 40:54
  SPEAKER_02: It's it's ludicrous.
• 40:56
  SPEAKER_02: Think about it.
• 40:56
  SPEAKER_03: Which part is ridiculous?
• 40:58
  SPEAKER_02: Do you think that do you think any judge or the American government is going to make open AI pay anyone$10 million?
• 41:05
  SPEAKER_03: I don't.
• 41:06
  SPEAKER_02: I don't either.
• 41:07
  SPEAKER_02: In fact, it's it's a non-starter.
• 41:09
  SPEAKER_03: I think the case will get dismissed.
• 41:10
  SPEAKER_02: But if Maryland Legal Aid did it, do you think they'd be made to pay?
• 41:15
  SPEAKER_02: Most definitely.
• 41:17
  SPEAKER_02: So yeah, I I see Raina's point because it is.
• 41:21
  SPEAKER_02: They are getting to do their trials.
• 41:23
  SPEAKER_02: They could have avoided all of this UPL issue by just not putting any legal information into their platforms, but all of them did.
• 41:32
  SPEAKER_02: They had, they knew we're giving it all of this information it's going to give out to the public.
• 41:38
  SPEAKER_02: They knew that was a risk when they were going in.
• 41:41
  SPEAKER_03: Well, you know what, Heather?
• 41:42
  SPEAKER_03: We're going to have a special show where you and I are going to talk about that documentary that we both we both just watched was.
• 41:51
  SPEAKER_03: I think we're going to do this for our 50th episode.
• 41:53
  SPEAKER_02: We both watched it twice.
• 41:54
  SPEAKER_02: We better watch it two more times before you get it.
• 41:58
  SPEAKER_02: What is it called?
• 41:59
  SPEAKER_02: I will send you a link, Arena.
• 42:00
  SPEAKER_03: I will send you a link.
• 42:01
  SPEAKER_03: It's called something like AI or how I become a couple of things.
• 42:04
  SPEAKER_02: It's called the AI doc or how I became an apocalyptimistic.
• 42:09
  SPEAKER_03: That's right.
• 42:10
  SPEAKER_03: That's right.
• 42:11
  SPEAKER_03: And there are some big there's some interesting things, but what you just talked about, Heather, was how they fed this legal information to the AI.
• 42:22
  SPEAKER_03: But you know what?
• 42:23
  SPEAKER_03: At this point, the AIs are smart enough to go out and find the information.
• 42:27
  SPEAKER_02: They're doing it on their own, but they did give it all the information and they didn't give it the guardrails to not give it out to the public.
• 42:33
  SPEAKER_02: So those those companies and the government already knew before this lawsuit ever came along.
• 42:40
  SPEAKER_02: So I don't have a whole lot of faith for what's going to happen, nor do I really think Nippian life is probably, you know, the the portion of the population that should be being really looked after.
• 42:51
  SPEAKER_02: I'm sorry, Nippian Life.
• 42:53
  SPEAKER_03: Well, you know what?
• 42:54
  SPEAKER_03: This opens up this opens up uh another segment.
• 42:59
  SPEAKER_03: Rita, we are in a users and builders season, which is really people are either going to say, give me a product off the shelf so that I can just, you know, give it my login information, I'll start using it.
• 43:12
  SPEAKER_03: Or they start saying, wait a second, I I want to I want to build something that works for my specific situation.
• 43:22
  SPEAKER_03: Which do you see be a more valuable approach for legal aid and access to justice organizations?
• 43:29
  SPEAKER_00: I I definitely think that we might be in the builder category.
• 43:34
  SPEAKER_00: I just looking at some of the tools that um Maryland Legal Aid has developed internally, I can't imagine that these are going to be off the shelf because they're so specific to the issues that these organizations face.
• 43:46
  SPEAKER_00: So, for example, one of the things that they have is like a rent ledger reviewer.
• 43:52
  SPEAKER_00: So one of uh this may be very niche, but if you go to rent court and you're an attorney representing people in rent court, one of the time consuming Tasks that you have is that you have to go through so many rent ledgers to find whether these fees are actually rent, were they something else?
• 44:06
  SPEAKER_00: Were they additional fees?
• 44:07
  SPEAKER_00: Are they even accurate?
• 44:09
  SPEAKER_00: And you have to spend time doing that.
• 44:12
  SPEAKER_00: And they developed some AI tool to review these ledgers for their attorneys so that they can, there's a time saving there.
• 44:19
  SPEAKER_00: They developed a brief uh creator so that it they could put in all the laws and all the statutes and case law related to certain topics.
• 44:29
  SPEAKER_00: So if they have to write a brief, they they actually have like a way to get a jump start.
• 44:34
  SPEAKER_00: And it's not like the whole brief, but at least gives you a first draft and and then you're able to review it.
• 44:39
  SPEAKER_00: So they're looking at very specific problems that their attorneys are constantly facing and that are time consuming and trying to solve for those.
• 44:49
  SPEAKER_00: And I can see what we came up with at the AI conference.
• 44:53
  SPEAKER_00: Yes, some things might be generic, but some things, if we can develop specific tools for these specific issues that our our community faces, and again, it's not like trying to do it in a way, what we learned from Maryland Legal Aid is that they've been doing it at pretty low cost and very organically because they have the experts to be able to do those things.
• 45:16
  SPEAKER_00: So they, if they can sort of do it for the community, I think that's really cool.
• 45:21
  SPEAKER_00: But I I don't discount, for example, at the I'm an employee of the Maryland State Bar Association.
• 45:27
  SPEAKER_00: We have a Google Suite, we have Google Gemini uh available to us and Notebook LM.
• 45:33
  SPEAKER_00: And I I love Notebook LM.
• 45:36
  SPEAKER_00: I I love it for research.
• 45:37
  SPEAKER_03: It's amazing.
• 45:38
  SPEAKER_00: I love it for organizing, I love it for summarizing, so many things.
• 45:43
  SPEAKER_00: There's so many ways to use it.
• 45:44
  SPEAKER_00: And I haven't probably only explored just the tip of the iceberg.
• 45:48
  SPEAKER_00: But those tools are really helpful and you don't need to build anything.
• 45:52
  SPEAKER_00: Like those tools are really effective and helpful.
• 45:54
  SPEAKER_00: So it's probably going to be and both things, um, but I see a use for each of them.
• 46:00
  SPEAKER_03: Well, that I mean, that's a great answer, and it actually leads us into our practice signal segment.
• 46:08
  SPEAKER_03: And this is a segment we have every episode where we're lawyers in the wild are grappling with a problem, and we're trying to see if we could help them with that problem.
• 46:18
  SPEAKER_03: And maybe AI can help them with the problem, maybe AI can't.
• 46:23
  SPEAKER_03: So, and we always get them from Reddit.
• 46:26
  SPEAKER_03: So a practicing lawyer posted online that they were trying to figure out how to address the access to justice problem in their community.
• 46:39
  SPEAKER_03: And they asked other lawyers three questions, which is where are the gaps in legal services?
• 46:46
  SPEAKER_03: Are there limited scope services lawyers could provide that would help fill those gaps?
• 46:52
  SPEAKER_03: And what are the primary barriers preventing people from getting appropriate legal counsel?
• 46:58
  SPEAKER_03: They even suggested something pretty specific, which is what if one lawyer handled the initial consult, identified the issues, and prepared a memo that another lawyer could then use to take the matter forward.
• 47:11
  SPEAKER_03: So when I read that today, I couldn't help but adding a fourth question is what parts of that gap can now be filled by AI?
• 47:20
  SPEAKER_00: Yeah.
• 47:20
  SPEAKER_00: And this is this is exactly the question that we've been trying to answer in this task force.
• 47:26
  SPEAKER_00: And so, for example, what we came up with in the task force as a as a recommendation is that there's a lot of expungements that need to be done in Maryland.
• 47:34
  SPEAKER_00: And it takes up a lot of resources and time for civil legal aid organizations to help clients fill out those forms and see that through.
• 47:43
  SPEAKER_00: And they can only serve a fraction of the clients that actually need expungements.
• 47:47
  SPEAKER_00: And expungements are important because if you have some conviction or even an arrest on your record, then you have all of these other barriers.
• 47:56
  SPEAKER_00: So you could have housing if you're trying to get housing, if you're trying to even get a job, if you're trying to get other public benefits, like that record will prevent you from doing that.
• 48:07
  SPEAKER_00: And so to me, when you have these high-volume cases, a lot of people need help.
• 48:12
  SPEAKER_00: And maybe it's a procedure that you can easily walk through, like a turbo tax.
• 48:17
  SPEAKER_00: What year did you get this?
• 48:18
  SPEAKER_00: And did you receive a null process or did you receive a STET or did you receive a conviction?
• 48:23
  SPEAKER_00: And it can walk through, walk you through all of those steps and come out with something where you can actually file it on your own because it walked you through something.
• 48:32
  SPEAKER_00: To me, that is one of the ways that legal technology can really help scale.
• 48:38
  SPEAKER_00: Because one of the problems in why do people not have the help that they need, we have about 40,000 lawyers in Maryland.
• 48:45
  SPEAKER_00: Probably 0.5% of those work in civil legal aid organizations.
• 48:50
  SPEAKER_00: So it's not that we don't have lawyers, it's just that the incentive structure is set up for lawyers to not serve people who are poor and meet this access to justice crisis.
• 49:00
  SPEAKER_00: So a couple of things we're doing.
• 49:01
  SPEAKER_00: One, you need to increase the supply of people who can provide legal advice and representation.
• 49:07
  SPEAKER_00: So there's more people in this space.
• 49:10
  SPEAKER_00: But no matter what, when you have people, there's going to be an upper limit to how many people they can serve.
• 49:16
  SPEAKER_00: And so I think what technology helps us break through is that you can scale at a level that we've never been able to scale at.
• 49:24
  SPEAKER_00: If we can do it in a safe way and if we can do it in a way that protects consumers, then I do think legal technology can be a huge solution here because then you can reach as many people, probably at very low cost at their own time, and they can start solving these problems on their own with limited legal human oversight, still having human oversight, but still very limited compared to having that case-by-case oversight.
• 49:50
  SPEAKER_03: Well, Rena, that was that was a very insightful response, and I really do appreciate that.
• 49:54
  SPEAKER_03: And it really looks like we've come to the end of a very interesting episode.
• 50:00
  SPEAKER_03: Thank you so much for being with us today.
• 50:03
  SPEAKER_00: Thank you so much for having me.
• 50:04
  SPEAKER_00: This was so much fun.
• 50:05
  SPEAKER_00: I really appreciate Heather for bringing this opportunity.
• 50:08
  SPEAKER_00: And Ron, it was fantastic talking to you and meeting you.
• 50:12
  SPEAKER_03: Great talking to you as well.
• 50:14
  SPEAKER_03: And this has been AI Tools for Practicing Lawyers.
• 50:18
  SPEAKER_03: We'll see you next time.
• 50:21
  SPEAKER_01: That's it for today's episode of AI Tools for Practicing Lawyers.
• 50:26
  SPEAKER_01: Thanks for listening.
• 50:27
  SPEAKER_01: We'll see you next time.