Key takeaways:

• • • Legal aid is leading on AI adoption 鈥� Legal aid organizations are leading the way in leveraging AI with 74% using AI in their work, driven by the need to serve millions of citizens who lack legal help.

    • Lone Star Legal Aid creates Juris 鈥� A new AI-powered tool Juris from Lone Star Legal Aid improves accuracy and trust through retrieval-augmented generation, source-cited answers, and a secure Azure-based architecture with an integrated citation viewer.

    • Keeping costs low 鈥� A phased, two-year build-and-test process kept costs low (at about $2,000 a year in infrastructure costs, plus about 300 staff hours) and produced dependable results.

A finds that under-resourced legal aid nonprofits are adopting AI at nearly twice the rate of the broader legal field because of the urgency of the need to serve millions of Americans who may lack legal help. The study shows that almost three-quarters (74%) of legal aid organizations already use AI in their work, compared with a 37% adoption rate for generative AI (GenAI) across the wider legal profession. (LSLA), a legal aid non-profit serving easter Texas, is one of early adopters of AI.

According to LSLA, its attorneys were spending too much time and money hunting for answers across pricey platforms and scattered PDFs. Key materials lived in research databases, internal drives, and static repositories, while individual worker-vetted documents were not centrally accessible. Without a single, trusted hub, staff experienced slower research time that affected clients through duplicated effort and delays.

These strains are not unique to LSLA. In fact, court help centers and self鈥慼elp portals face the same fragmentation, licensing costs, and uneven access to authoritative guidance. A verifiable, consolidated knowledge hub that could stabilize quality while reducing spending would be a needed solution.

To solve this problem, LSLA turned to AI to create a legal tool called Juris built to return fast, source鈥慶ited answers. Juris was designed to centralize high鈥憊alue legal materials, cut reliance on expensive third鈥憄arty platforms, and lay a flexible foundation that the organization could reuse beyond legal research for internal operations and future client tools.

Multifaceted approach to ensuring accuracy and reliability

There were several aspects of Juris that designers used to help its mission to increase access to justice, including:

Design methods fuel trustworthy output 鈥� Juris was built to ensure accuracy using a number of methods, such as a retrieval-augmented generation (RAG) pipeline to ensure the chatbot delivers fact-based, source-cited answers. It also uses semantic chunking, a process that breaks a document into natural, meaning鈥慴ased sections (for example, a heading plus the paragraphs that belong to it) so the original context stays together.

When a user asks a question, Juris retrieves only the most relevant of these sections. Limiting the AI to evidence from those passages improves accuracy and reduces hallucinations because the model is not guessing from memory. Instead, it is grounding answers in the text it just accessed.

Solid technical architecture helps reliability 鈥� Juris鈥檚 technical architecture also ensures reliable results because it combines Azure OpenAI for secure, stateless access to AI models to better handle document ingestion, processing, and vector storage. Users interact through a custom internal web interface that integrates a PDF viewer alongside the chat experience that enables seamless citation and document navigation. The platform is securely hosted on Azure App Service with continuous deployment orchestrated through GitHub, which provides reliable operations and streamlined updates.

Phased approach to building and testing yielded dependability 鈥� Also to ensure trustworthy results, LSLA developed Juris by following a structured, phased approach over two years. It began with a concept phase that was focused on clearly identifying the problem, followed by a platform evaluation that compared open-source and commercial solutions. A prototype was then created and demonstrated as proof of concept.

In addition, internal testing included adversarial exercises, hallucination detection, and rigorous validation of citation reliability. Based on these findings, the team implemented enhancements, such as moving from size-based to semantic chunking, improving the interface, and expanding the set of source materials. Juris is now in pilot preparation and undergoing final refinements before its release to a select group of subject matter experts.

Efficient resourcing and sharing learnings

LSLA鈥檚 phased method to building and testing also made sure that sustainability was built in from the beginning. Indeed, ongoing maintenance is minimal, and Microsoft鈥檚 nonprofit Azure credits keep infrastructure costs around $2,000 per year.

The most significant cost was in staff time. Development so far totals roughly 300 staff hours (or about 0.5 full-time equivalent, plus 0.3 FTE over two years). Once Juris enters phase two, which has been funded by a Legal Services Corporation (LSC) technology initiative grant, expected benefits will include faster, more consistent research and reduced workload for frontline and administrative staff, plus a modular framework that others can adapt.

Other legal service organizations that face similar challenges can learn from the Juris development, testing, and implementation as well as other related case studies. These recurring lessons include:

• • • beginning with a small, manageable scope
    • inviting end users in from the start, and
    • carving out protected time so staff can innovate alongside daily duties.

Looking ahead, the LSLA team will continue to roll Juris out in phases, while building sister tools. LSLA also plans to share lessons learned through LSC鈥檚 AI Peer Learning Labs to help other organizations replicate the model.

Real change at scale, such as this, will only come from collaborating across organizations to share playbooks, pool datasets, and co鈥慸esign tools that lift quality while lowering cost. It is only with such partnership and sharing lessons from early adopters of AI that peers can adapt the model and, together, scale solutions that narrow the justice gap.

Angela Tripp, Program Officer for Technology for the Legal Services Corporation contributed to this article.

You can learn more about the ways legal aid organizations are using advanced technology to better serve individuals as they access the justice system here

Now, as AI-driven tools and technologies rapidly evolve, they offer unprecedented opportunities to scale solutions and bridge this divide through automation, information retrieval, and process optimization.

However, the promise of AI in reducing the justice gap lies not in replacing human connection but in enhancing it. While technology can expand reach and efficiency, the human elements of empathy, trust, and contextual understanding remain irreplaceable in legal assistance.

, Executive Director of (LLL), says she this fact over her eight years of working at the intersection of the justice gap and technology. By combining AI with human-centered design principles, Brown and her peers are providing efficient referrals to enable legal aid organizations to connect with humans faster.

Leveraging AI to expand human capacity

The most effective AI implementations recognize a fundamental truth: People in legal crisis crave human connection. “We hear constant feedback on our tools along the lines of ‘I just want to talk to someone,鈥欌�� says Brown, adding that this insight is driving a shift in focus in LLL鈥檚 work toward answering how the organization might best maximize meaningful human interaction between potential clients and legal service providers.

Across Louisiana, Brown is undertaking innovative projects that seek to expand awareness around legal issues, open access to self-resolution pathways, and provide more sophisticated referrals, working toward a future-state in which there are more high-quality interactions between service providers and those seeking legal help.

Register now for The Emerging Technology and Generative AI Forum, a cutting-edge conference that will explore the latest advancements in GenAI and their potential to revolutionize legal and tax practices

At the current moment, Brown acknowledges that there is a lot of work required to leverage technology and reduce administrative burdens in order to pave the way for a future in which human-to-human interactions are paramount. LLL is is doing its part by using AI as a force multiplier for existing services, including:

Creating civil justice content 鈥� LLL is using AI to accelerate the production of written articles, video scripts, brochures, translations, and social media content that can be used to help build awareness of tools and resources available for civil justice needs.

Offering sophisticated referral systems 鈥� LLL is developing an AI-powered referral function that can take complicated intake rules and case acceptance guidelines from Louisiana civil justice organizations and match those seeking legal help with the appropriate resources. The organization is also building AI-supported content-retrieval functionality within its web site that will issue relevant spot-and-select information guides, self-help tools, and referral resources from the organization’s database.

Modernizing traditional processes 鈥� Brown and her team are working on another innovation: Paper-to-digital intake workflows supported by AI, which can reduce the time needed to input data and leave more time for a conversation with a potential client.听Less tech-savvy individuals or those that have lower access or awareness of resources still will be able to use a paper-first approach to connect with service providers that can help.

AI tools are beneficial, but a cohesive strategy is essential

The true potential of AI tools to maximize human-centered approaches in legal services can only be realized when such approaches are incorporated into a comprehensive and unified strategy to close the access to justice gap for good, explains Brown. Right now, this is lacking, she says, adding that we need a strategy 鈥渢hat firmly orients our community around a broad vision with specific outcomes and goals, but still encourages experimentation and innovation from a place of clarity about where we are going. We ultimately have to get very clear on who’s doing what, what capacities we have or don’t have, all the various inputs that make up the system as a whole,听and importantly, how all of it relates to each other.鈥�

In addition, the assessment process of gaining clarity regarding goals, outcomes, and existing capacities is a key requirement for legal aid鈥檚 success in deploying AI tools with empathy-driven design. The most effective implementations are built upon established processes and use AI to enhance rather than replace existing frameworks. And Brown鈥檚 initiative to transform paper intake forms into hybrid workflows while preserving familiar client experiences is a notable example.

Until an integrated approach is realized, Brown and her team at LLL will keep working to embrace technological advancements to better streamline administrative processes and foster a future that emphasizes the importance of close human interactions. This reflects her current ethos around AI 鈥� and around technology in general 鈥� that lawyers and access to justice professionals should be working to maximize human interaction wherever possible.

鈥淵ou cannot replicate compassion and understanding for those in crisis with a decision tree or a chatbot,鈥� she says, noting that AI can expand our capacity in ways that leave time for these crucial, personal interactions, and allow us to 鈥渢houghtfully design experiences that make our humanity the centerpiece in a very highly automated service.鈥�

You can learn more ways that AI-driven technology is impacting access to justice here

Pursuing a career in law with a focus on public service inherently aligns with the desire to contribute positively to society. Indeed, public service law offers attorneys a profound sense of purpose and fulfillment, a key ingredient for the well-being of lawyers.

Yet first-year law students find out quickly that their prime reason for entering the legal profession can be drilled out of them by the realities of law school. One of these is known as the law warrior culture, a term describing the profession鈥檚 premium on and showing a minimum of vulnerability. Another reality is the high cost of law school, which leaves students who entered the profession for altruistic reasons finding themselves having to postpone potential dream jobs and instead seek high-paying jobs to pay down their large student loan balances.

Register now for The Emerging Technology and Generative AI Forum, a cutting-edge conference that will explore the latest advancements in GenAI and their potential to revolutionize legal and tax practices

Both of these factors contribute to the ongoing access to justice gap and the high cost of legal representation. Most low-income Americans do not get any or enough legal help for their civil legal problems, and the cost of legal help stands out as an important barrier, according to the .

Indeed, this broken social contract between the legal profession and the public is something that could be improved by more widespread us of AI, says , Esq., a corporate legal strategist who works with founders and CEOs. Clark鈥檚 decade-long work and leadership on lawyer well-being gave her insight into how AI could significantly improve lawyer well-being by radically changing how we value legal services.

The question hinges on whether AI could both improve access to justice and attorney well-being; and Clark and , Legal Futurist at Filevine, believe so.

Ways AI helps improve junior lawyer well-being

AI emerges as a transformative force by offering significant opportunities to enhance lawyer well-being, especially for junior lawyers, according to Boyko, who has used her law degree to push innovation in law firms, especially around issues of听legal tech, in-house talent, and legal education. Her interests 鈥� which converge at the intersection of technology, the business of law, law students and legal careers, and professional well-being 鈥� have served as a focal point for her observations about AI and junior lawyers.

Specifically, Boyko highlights the potential of AI in streamlining low-level tasks, which often burden young attorneys. Freeing junior lawyers up from this allows them to focus on more meaningful work. 鈥淚 like to think about [using AI] in terms of being able to minimize or streamline a lot of the rote low-level tasks that lawyers often find themselves doing, which I think are a huge drag on our well-being as attorneys,鈥� she explains.

Boyko says she also sees AI as a way to nurture curiosity and proactive learning. For example, AI can provide context and resources to junior lawyers, which is something that senior lawyers too often lack the time to do because of their own billable hour demands.

Finally, AI can benefit junior lawyers during times when they feel uncertain and unsure of the direction to take. 鈥淎I can help you think through situations where you may not have answers,鈥� she notes. 鈥淏eing able to use AI鈥� can really help to support well-being when you feel very lost and there’s a lot of pressure to find the answer quickly.”

Shifting to AI-driven value-based services

Clark advocates for the use of AI to help lawyers鈥� well-being in an industry that鈥檚 rapidly shifting from time-based to value-based metrics to better track performance.

鈥淲hat does it mean to have value outside of your time output?鈥� Clark asks. 鈥淚 think our value is judgment, discernment, and experience 鈥� things that just are not measured by time.” Requiring a quota of billable hours as part of the law warrior culture is often a significant source of stress and burn-out for lawyers. Value-based billing instead offers lawyers the opportunity to be paid for their judgement, expertise, and experience as a source of value rather than by hourly billings.

And AI can enable this shift away from the traditional billable hour to a value-based model as well as an overall transition from a reactionary model to a proactive type of workflow in order to improve client service, Clark says, drawing a compelling analogy between initiative-taking lawyering and preventive healthcare. Just as preventive care can avert serious health issues, proactive legal advice can prevent costly legal problems, she says. “We no longer practice or incentivize proactive lawyering,鈥� Clark adds. 鈥淲e’re paid to be reactionary.鈥�

This shift to value-based service allows legal professionals to reclaim a portion of their time and dedicate it to proactive lawyering. In this way, AI can become a thought partner in legal work, empowering lawyers to engage in deeper, more nuanced legal analysis and client service, Boyko explains. And as stated earlier, more proactive legal service is a top motivator for lawyers and source of well-being 鈥� both of which are key ingredients for lawyers to achieve their professional purpose.

AI planting the seeds for change

To foster positive well-being among lawyers, it is crucial for legal organization leaders to address the challenge of overcoming fear and resistance to change. AI has the opportunity to enable this paradigm shift by cultivating a culture of openness and continuous learning within the legal profession. And by encouraging curiosity and providing training on AI tools, organizations can help their lawyers become more comfortable with new technologies.

The integration of AI into legal workflows will transform attorney well-being and help avert the impact of several additional headwinds facing the profession, including the entrenched billable hour model, reaction-based lawyering, and the gap in access to justice.

In fact, some believe the question is not if AI will spur these changes, but when? The legal profession stands on the brink of a transformative era, Boyko and Clark say, and it鈥檚 one in which enhanced well-being, client service, and lawyers鈥� futures as well as the communities they serve all can thrive.

You can find more about the use of AI and GenAI in the legal industry here

For example, the developed the chatbot Beagle+ to assist people with step-by-step guidance on everyday legal problems. , Digital & Content Lead at the People鈥檚 Law School, led the efforts to create Beagle+ with technical assistance from听, Founder of Tangowork. And the Alaska Court System (ACS) and听 using a grant from the NCSC to develop an AI-powered chatbot called the听Alaska Virtual Assistant, or AVA.听Jeannie Sato, Director of Access to Justice Services of ACS worked with , CEO and Founder of LawDroid, to develop the tool.

How courts can successfully experiment with AI

Jackson, McGrath, Sato, and Martin all offered their step-by-step guidance on how courts and nonprofits can experiment and use AI successfully within courts systems.

Step 1: Determine the problem

When starting a generative AI (GenAI) legal assistance project, it is crucial to first pinpoint the specific legal needs and challenges faced by your target audience. McGrath noted that he sees several common examples, including providing public access to legal information, creating internal resources like bench books for judges, and automating court document preparation.

To properly identify the problem, conduct thorough user research to understand pain points related to accessing and applying legal information. For instance, Martin suggests starting by speaking with court staff. “I think we sometimes get caught up in the excitement about wanting to throw AI at the problem and create a solution,鈥� Martin explains. 鈥淎nd there are many use cases, but I think the part that’s really important is to meet with your staff, meet with everyone who’s being impacted by the burden of work, and then determine, based on that, what is the best choice.”

Taking the time upfront to clearly define the problem will help ensure that any AI solution being developed is truly meeting a demonstrated need.

Step 2: Craft a vision

Shifting from problem identification to crafting a vision for the GenAI-powered solution is crucial. The People鈥檚 Law School鈥檚 Beagle+ chatbot illustrated this well. 鈥淏egin with the end in mind,鈥� says Jackson. 鈥淲hen you begin a project, keep in mind what you’re trying to achieve and what success looks like because that’s going to be different for each person.鈥�

Jackson further described how in 2018, the initial vision was to create a chatbot capable of intelligently answering questions about consumer and debt law in British Columbia. Today, while that vision is realized, the ability of GenAI technology to adapt and improve over time necessitates a continuous and evolving vision.

Step 3: Allocate realistic resources

Assessing available resources is crucial before embarking on a GenAI project, with a realistic evaluation considering such factors as existing legal content, technological capabilities, staff expertise and capacity, and budget.

It鈥檚 important to examine the state of the organization鈥檚 existing legal information, including its documents and web pages, to determine the quality and consistency. Indeed, conflicting information across sources often can confuse GenAI models.

For staff capacity, Sato explains how the ACS started with a small team of people, which included the court administrator, the chief technology officer, a webmaster, and two to three staff attorneys, who were necessary for content review, testing, and feedback. It is not uncommon for an initial project to consume about 30% of each team member鈥檚 time.

Technological expertise is also a key consideration in resource assessment. In fact, Martins says this underscores the importance of working with a technology partner that can help navigate the different choices and options available, including the need to understand options for AI model selection, vector databases, and embedding strategies. While some may consider using large language models (LLMs)to reduce costs, the expenses for setup and maintenance often outweigh the benefits compared to using established services like OpenAI.

Financial resources are also a consideration, of course; however, it is worth noting that the cost of OpenAI tokens is often surprisingly low compared to other project expenses. For the creators of Beagle+, for example, using OpenAI鈥檚 tool has cost no more than $75 per month, according to Tangowork鈥檚 McGrath.

Courts can explore the possibilities of AI tools in tackling their specific legal challenges by experimenting within

Addressing common concerns

Our experts say that two common concerns often arise when considering the use of GenAI to solve justice gaps: one is the need for multilingual capabilities; and the second is how to handle AI-generated inaccurate information, or so-called hallucinations.

鈥淎dvanced LLMs like GPT-4 demonstrate impressive multilingual capabilities and are able to understand and respond in numerous languages on-the-fly without requiring additional training or configuration,鈥� explains McGrath. 鈥淢ultilingual support is a key advantage of modern LLMs, enabling chatbots to serve diverse populations with minimal additional development effort.鈥�

However, hallucinations are a significant concern when using LLMs for legal applications. Fortunately, the combination of several advanced strategies can mitigate hallucinations:

• • • First, grounding responses in providing context through techniques like retrieval-augmented generation can help tether outputs to verified source material.
    • Second, careful prompt engineering and relevancy scoring can further constrain responses.
    • And finally, automated checks that compare model outputs to source documents can flag potential hallucinations.

At the same time, manual expert review by humans 鈥� known colloquially as human in the loop 鈥� remains crucial, even with automated safeguards in place. Therefore it is key to periodically sample responses for human verification and focus more intensive review on higher-risk conversations.

Creating a successful AI-powered chatbot for legal information requires careful consideration of the several steps cited above. By following these actions and staying up to date with the latest developments in AI technology, courts and organizations working to close the justice gap can create effective and responsible chatbots that provide valuable legal information to those who need it most.

You can register here for the upcoming NCSC webinar on March 19, which will explore the

With more than 20,000 children in Florida鈥檚 foster care system at any given time, Sartor is setting a precedent for how technology can be harnessed to inform, guide, and inspire youth towards a better future.

From advocacy to action

Sartor鈥檚 journey began with a during college, in which she volunteered to advocate for children in foster care. Her passion deepened during her time as an AmeriCorps member, mentoring students in a pilot program. While at AmeriCorps, Sartor represented one of her students, improving his life while solidifying her interest in attending law school to represent even more youth with urgent legal needs.

Sartor鈥檚 dedication carried through to law school, where she earned an Equal Justice Works Fellowship at the Children鈥檚 Law Center. One specific challenge continued to stand out: the absence of a centralized resource in Florida to help foster youth understand their rights and access critical services. Inspired by initiatives in other states, Sartor partnered with law students to create a know-your-rights guide, which she used in her legal practice. The project gained even greater traction after a grant from the allowed the guide to be digitized.

FosterPower was developed with a clear vision: to inform foster youth about their benefits, protections, and legal rights. 鈥淲e don’t cater to any other audience than the youth 鈥� FosterPower is made for them,鈥� Sartor explains. 鈥淗owever, the content is so simplified and cites to the law at the bottom of each section that it is truly a useful resource for anyone in child welfare. We do have case managers, attorneys, and judges that use the app, but our focus audience has been and always will be youth in foster care.鈥�

To build the app, Sartor took a hands-on and sensitive approach by working directly with foster youth. 鈥淭hese youth have experienced significant trauma and talking about their experiences can be triggering, so I wanted to make sure that it was approached in a youth-centered, trauma-informed manner,鈥� she says. 鈥淚 did not want to outsource this task to a vendor that did not have the experience working with this vulnerable population.鈥�

鈥淲e don’t cater to any other audience than the youth 鈥� FosterPower is made for them. 鈥淗owever, the content is so simplified and cites to the law at the bottom of each section that it is truly a useful resource for anyone in child welfare.”

 

Sartor notes that the biggest factor was really just listening to the youth and creating an app that is based on their feedback and what they wanted to see when they pulled up FosterPower. She also says they prioritized compensating foster youth for their time, ensuring that they were truly valued and respected as a part of the process.

The result is an app that鈥檚 as practical as it is accessible. Available on both Android and iPhone 鈥� a must-have to Sartor 鈥� FosterPower was designed to function offline, ensuring that users can access critical information anytime, anywhere. Content is reviewed on an annual basis with local experts, ensuring that information is as up to date as possible.

A vision for the future

Since its launch, thanks to word of mouth, local partnerships, and social media posts, FosterPower has achieved impressive milestones. The app has been downloaded more than 4,000 times, and its website has gathered more than 10,000 views. The app鈥檚 educational videos have reached more than 100,000 viewers 鈥� proof of the demand for foster care-related education in Florida in beyond.

To further its reach, Sartor and her team conduct in-person presentations and CLE sessions. Looking to the future, Sartor says they鈥檒l be hiring a community marketer and trainer to build relationships with case managers and ensure wider adoption of the app. Recent updates include an immigration section available in Spanish and Creole, as well as a forthcoming human-trafficking module, reflecting the diverse needs of Florida鈥檚 foster youth and FosterPower鈥檚 commitment to addressing them.

Sartor鈥檚 current focus is on ensuring all foster youth across the state are empowered with the information they need to succeed. 鈥淲e continue to work on getting it out across the state so that all youth know about it,鈥� she explains, adding, however, that her ambitions for the app extend beyond Florida. With the help of funders and other local experts, she says she envisions expanding FosterPower to other states, encouraging local organizations to adopt and adapt the platform for their own unique needs.

For those inspired by Sartor鈥檚 work, her advice is clear: Listen to the population you aim to serve, involve them in the design process, and compensate them fairly. She stresses that building tech solutions for vulnerable populations requires deep empathy, adaptation, and a commitment to sustainability.

FosterPower is a unique opportunity to change lives for some of the country鈥檚 most vulnerable 鈥� foster youth 鈥� and the reception of their app is a testament to their its thoughtful design and development process, which stands as a testament to the power of user-centered design and compassionate innovation. With every download, FosterPower moves closer to ensuring that no child in foster care is left without the resources they need to succeed and thrive.

You can find out more about how legal technology is helping further the cause of justice here.

From complex analysis to AI-driven legal research, here are two compelling examples of how AI is already helping LSOs enhance their work.

Case 1 鈥� The California Innocence Project: Accelerating case reviews

The California Innocence Project, dedicated to exonerating wrongfully convicted individuals, was accustomed to reviewing thousands of files, with each typically around 50-plus pages, by hand, each year to determine which cases to pursue. The manpower required to complete such a massive manual review processes was unwieldly and inefficient. So, when the Project鈥檚 former Managing Attorney Michael Semanchik was introduced to proprietary AI-powered suite of legal tools at a conference, he decided to give it a try.

鈥淲e started testing it weekly, and I soon realized it was more powerful than I had anticipated,鈥� says Semanchik. With the new tool, the organization could upload lengthy case files into the system, and the tool would outline the files鈥� contents. It also would respond to specific, complex questions 鈥� such as, 鈥淲hat was the age of the defendant at the time and was there any evidence of low IQ?鈥� The tool also could flag any inconsistencies in testimony.

In one early case, the team had an arson hearing and asked the tool for a potential line of questioning specific to a government official. 鈥淚t perfectly gave me that set of questions,鈥� Semanchik says. 鈥淚t wasn鈥檛 an A+ because it wasn鈥檛 case specific, but it got all of the basics.鈥� Of course, human oversight is still crucial to ensure accuracy and reliability, he says, adding that attorneys 鈥渟hould double-check every response you get.鈥�

Overall, Semanchik believes in the power of AI to assist with the Project鈥檚 work. As a next step at his new organization, The Innocence Center, he says he plans to incorporate a baseline set of complex questions for the AI tool to expedite his review. At scale, he鈥檚 excited about the potential to help self-represented litigants directly.

鈥淚magine that litigant sitting in prison who can use AI to create a statement of facts for their Habeas corpus or draft applicable claims in a better way,鈥� he explains, adding that AI-driven tools like this will give defendants a better chance at justice. 鈥淚t opens up doors for people without a legal degree to access justice from inside prison. It鈥檚 unheard of 鈥� we will get to the truth faster than ever before.鈥�

Case 2 鈥� Housing Court Answers: Empowering staff & tenants with AI-driven legal tools

In New York City, Housing Court Answers (HCA), a tenants’ rights organization, teamed up with a different legal automation platform along with Sateesh Nori, an Adjunct Professor at New York University, and others at Cornell University, to create two AI-driven tools designed to assist tenants facing eviction or housing instability. The first of the tools is for internal use by HCA staff to provide guidance to tenants and draws on academic legal resources. The second, which is more limited in scope to reduce potential risks to users, provides tenants with answers to basic housing questions through the HCA website.

Responding to tenants鈥� legal concerns is a 鈥渉igh pressure situation,鈥� says one executive involved in the HCA project. 鈥淪taff can find it difficult to access and properly use complex resources when they鈥檙e answering phones or attending one of the Help Desks, so by giving them access to extensive materials in an accessible way, they can simply enter the question into their app and get guidance.鈥� Having a self-serve FAQ tool on their website, while not able to address all situations, can help ease volume and provide tenants with resources.

The teams turned to large language models (LLMs) as a solution because they knew they had a myriad of answers documented, but 鈥渘eeded a way of surfacing it that was accessible, helpful, and safe,鈥� says the executive. 鈥淭his is one of the things that a closed-domain generative AI-powered tool does extremely well: find the relevant information, and surface it in an easy-to-digest way.鈥�

To build out the application, the teams collaborated on four main steps. First, they compiled the corpus, or knowledge base, by gathering existing resources, expert input, and student research into a clean data set. Second, they tested and trained the tool. The subject matter experts used the platform鈥檚 in-built human-in-the-loop moderation functionality 鈥� in which humans intervene in computer-generated answers to ensure quality 鈥� in order to improve the performance of the tool to an acceptable level. Third, they launched the app internally and externally, on HCA鈥檚 website, to engage with users. Finally, they 鈥渃ompounded learnings from the data, analytics, and generated responses to both improve the tools and to understand the kinds of issues impacting tenants at any given time,鈥� explains the executive. 鈥淭his data is gold for advocacy efforts.鈥�

Key to the HCA project鈥檚 success has been keeping people at the center of AI-driven solutions from human-in-the-loop development to continuous testing to iteration. 鈥淗ousing Court Answers staff members are very excited about the potential of AI in our work helping tenants facing eviction and housing instability,鈥� says Jenny Laurie, Executive Director of Housing Court Answers. 鈥淲e are looking forward to the stage in which the AI tool can answer many of the basic questions people have about their rights so that our human staff can tackle the more complicated requests for help 鈥� and to the stage in which the AI鈥檚 knowledge base can help train our new staff, because the rent and eviction law and regulations in New York are incredibly complicated and hard to digest.鈥�

This initiative demonstrates the potential of legal aid organizations to leverage existing resources for both internal and external use, thus saving valuable staff time.

Continuing to leverage AI for legal aid

These two case studies highlight the power of AI to not only strengthen legal aid organizations鈥� work internally to allow them to serve more people, but also to provide critical resources to individuals seeking legal information on their own. By taking a human-centered approach to design, double-checking outputs, and continuing to improve the AI models, these pioneering groups are creating playbooks that over time, will significantly change how individuals increase access to justice at scale and at a much lower cost and with a greatly reduced time commitment.

In this two-part blog series on , we will look at how AI-driven technologies can help those access legal aid to better secure results. In the next installment, we will look at how AI can empower those clients most in need.

Kristen Sonday: Tell us about your role within the Legal Services Corporation.

Holly Stevens: Within LSC, I lead the Office of Data Governance and Analysis. We are a group of data professionals 鈥� researchers, data engineers, data scientists, analysts, and web developers 鈥� working to provide data to legal aid organizations and other stakeholders.

Kristen Sonday: What is the Civil Court Data Initiative?

Holly Stevens: Every day, millions of Americans struggle with civil legal issues that limit their access to housing, healthcare, employment, education, and more. When these cases go to court, they are heard in more than 15,000 individual state and local courts across the country, and most individuals do not have access to legal assistance or representation.

Due to the highly decentralized nature of our courts, key stakeholders lack access to data about these cases 鈥� for example, the impact of legal representation (or lack thereof) and the case outcomes. This lack of access to data hinders transparency, informed decision-making, resource allocation, the identification of disparate treatment and outcomes, and the ability to measure the impact of interventions. Without data, it’s challenging to address systemic issues, improve access to justice, and ensure that the court systems operate in a fair and equitable manner.

The CCDI began in 2019 to determine how access to real-time civil court data could inform legal aid providers鈥� response to changing local needs. Our vision is to democratize court data and get it in the hands of organizations who can help enhance engagement with the courts, decrease default or failure-to-appear rates, and improve legal services use of data.

Kristen Sonday: How does CCDI work, and what are a few examples of real-world applications for the open-source library, CleanCourt?

Holly Stevens: In the existing system, the names of those involved in a civil case are recorded in electronic filing systems as free text or transcribed from paper dockets, leading to typos or differences in formatting when recorded over time. While these differences may be slight, they can pose significant challenges when analyzing the data in aggregate 鈥� for example, when a corporation has changed its business designation multiple times.

CCDI has developed several processes for cleaning names in aggregate from court lookup websites and data-sharing initiatives and has consolidated these methods in the . These methods employ natural language processing to parse and group party names and are more complex than common methods of data cleaning. As a result, we can drastically reduce the time required to run the complex string cleaning with high-quality results. In fact, CleanCourt can analyze more than four million records in less than 90 minutes.

Standardizing court data is important so we can have high confidence and integrity in the data with which we鈥檙e working and understand the impact of certain patterns on the justice system accurately. For example, , showed that more than 70% of debt collection cases were filed by just 10 companies. The same trend has been documented in eviction cases in cities across the country 鈥� a small number of landlords disproportionately drive eviction filings. Being able to identify the repeat players who commit bad practices, overload our court system, and treat tenants unlawfully is crucial to ensuring these cases are properly addressed.

In another example, as pandemic-related eviction protections expired in Virginia, we developed an early warning system for eviction filings accessible to legal aid providers across the state. The system includes weekly reports on evictions in every local jurisdiction in Virginia. Advocates use the reporting to develop outreach activities, including having lawyers in courts on specific dates that have a large number of eviction hearings scheduled, and working with community organizers to respond to mass evictions in specific properties.

Kristen Sonday: What are some of the challenges of using this data for good?

Holly Stevens: The biggest challenge is balancing the privacy of individuals with the need for data collection and increased transparency. There鈥檚 a real risk for people facing eviction or debt collection in which commercial entities can use the data to create better algorithms to screen tenants with a prior eviction or use predatory lending schemes with those who have faced debt collection.

Kristen Sonday: How can other organizations access LSC鈥檚 civil court data and collaborate?

Holly Stevens: Organizations can access the data set at , and we are always open to collaboration. We work directly with legal aid organizations to customize data for their use cases, which is important given their lack of resources. We ensure we get them what they need to improve outreach, put lawyers in rural courts when hearings are scheduled, evaluate their practices, or advocate for systemic change.

Kristen Sonday: How do you think about CCDI鈥檚 potential impact?

Holly Stevens: When you鈥檙e looking at millions of records of data that are messy and challenging to understand, you have to remember that this data is really about people. People who are parties to a case, people trying to keep up with fast-moving dockets, and people who are trying to help (and some who aren鈥檛).

Our biases are reflected in the ways we鈥檝e collected the information, the ways we have or have not invested in better data collection and analysis, and the ways we analyze the data. Courts are in the unenviable position of handling these cases which often the law is ill-equipped to help resolve. Low-income families may have one triggering event that led to the legal problem 鈥� such as a medical bill or a child in the hospital 鈥� that can lead to job loss and then missed rent payments. We see so much opportunity in this data to help.