On a recent Monday, seven teams of law students sat down with laptops, picked a legal scenario, and started building. No one had prepared in advance. Most had never used AI tools at this scale—building full workflows and prototypes from scratch. By the end of the night, multiple teams had produced working legal technology prototypes, including one with document management, automated issue spotting, contradiction detection, and a first-week action plan generator.
That team, Kaitlin and Andre, are first-year evening students. They have full-time jobs. They had never built anything like this before. And in about two hours, they created something that, with continued iteration and real-world feedback, could see use in a small firm setting.
This was LITSA’s inaugural Legal AI Innovation Challenge, and what happened in that room confirmed something we had suspected but needed to see: law students are natural builders, and the tools have finally caught up to their instincts.
The scenarios
We designed two tracks, each grounded in the kind of work that legal professionals face every day.
In Track A, participants stepped into the role of intake coordinators at a legal aid organization facing a surge of client requests. They received a stack of intake narratives from multiple clients, each with different legal issues, varying levels of urgency, and incomplete information. Some clients were facing eviction hearings in days. Others had wage theft claims with approaching statutes of limitations. One client’s narrative raised potential domestic violence concerns buried beneath a housing complaint. The challenge was not just identifying what each person needed, but triaging under pressure, flagging gaps, and drafting clear, empathetic client communications that a real person in crisis could actually understand.
In Track B, participants played a junior associate at a small firm. A new client had walked in with a commercial lease dispute involving allegations of retaliation, potential building code violations, and a landlord threatening eviction. The fact packet included the lease agreement, correspondence between the parties, a notice of default, and a client intake summary. Students had to assess the claims, identify defenses, spot what was missing, and produce something a supervising attorney could pick up and use the next morning.
Both tracks were deliberately messy. Real legal work is messy. We wanted to see how students navigated ambiguity, not just how well they could prompt.
What they built
The range was striking. Some teams used ChatGPT. Others used Claude. A few used code editors and built interactive web applications. None of them had done this before that evening.
Kaitlin and Andre’s winning submission, LegalTree HAI, went well beyond what we expected from a first attempt. They built an MVP for a small-firm intake coordination platform with document management, AI-powered fact verification that flagged contradictions across documents, automated issue spotting, gap analysis, and a first-week action plan generator. As with any prototype built in two hours, continued development and closer inspection would surface plenty to refine, but the core logic was sound and the workflow was genuinely usable.
Another team built an interactive early case assessment application using Claude that identified Massachusetts Home Improvement Contractor Act claims with potential treble damages. They structured the analysis to walk through elements and defenses methodically.
One student used ChatGPT to systematically triage four client matters and found that the AI surfaced ambiguities in the intake narratives that she had initially missed on her own read-through. The tool did not replace her judgment. It made it sharper.
Another student produced a comprehensive seven-section triage memorandum with a reusable decision framework, essentially building a template that a legal aid organization could adapt and deploy across future intake cycles.
View a summary of the Legal AI Innovation Challenge results (PDF).
The problems worth solving
According to the Legal Services Corporation’s 2022 Justice Gap Report, LSC-funded legal aid organizations must turn away one out of every two people who come to them eligible for help, simply because there aren’t enough resources to go around. That turn-away rate has not improved since LSC first measured it in 2005. In state civil courts, more than 75% of cases involve at least one party with no legal representation at all.
The traditional response to this has been admirable but structurally limited. Legal aid clinics, pro bono hours, one-to-one representation. These matter enormously, and they will continue to matter. But they do not scale. There are not enough lawyers, not enough hours, and not enough funding to close a gap that wide through individual service alone.
What scales is technology. And building that technology is now dramatically more accessible than it used to be. The inference costs of running AI models (the per-query cost of getting a response from a large language model) have dropped to fractions of a cent for many use cases, making it cost-effective to build and operate platforms that would have required a full engineering team and significant capital just a few years ago. The range of what a single motivated person can build has widened considerably.
The problems worth tackling go well beyond intake triage: document review at volume, knowledge management across cases, compliance monitoring, client communication in plain language across multiple languages, navigating benefits applications and housing disputes and family court procedures. These are structured, text-heavy, rules-based problems. Current AI tools are already more than capable of helping with them, and have been for months. And law students bring a depth of understanding of these problems that is hard to find anywhere else.
The complexity of building reliable legal technology is real. But the people best positioned to identify what needs to be built, and to evaluate whether it works, are the people who understand the law.
The lawyer as builder
What we saw in that room was not engineering. It was legal problem-solving with new tools. The students who produced the strongest work were not the most technical. They were the most curious. They asked better questions. They explored options and followed threads. They refined their approach when the first output fell short. They understood the problem deeply enough to recognize when the AI was helping and when it was drifting.
Law students are trained to read carefully, think structurally, and argue precisely. They identify issues, weigh competing considerations, and construct frameworks for decision-making. These are the capacities that make someone effective with AI tools. Not typing speed or programming knowledge, but the ability to articulate what you need, evaluate what you get back, and iterate until it is right.
And the tools are improving fast. What students built in March 2026 with generally available AI tools would have required a dedicated engineering team and months of development time just a few years ago. The capabilities today may look modest compared to what will be possible by the end of this year. These models are getting measurably better with each release cycle, and the gap between “I have an idea” and “I have a working prototype” keeps shrinking. That has real implications for who gets to build legal technology and how quickly problems get addressed.
This is already happening at Suffolk. 2L Akila Narayanan built FamilyShield, a platform addressing family law challenges, and recently won Hofstra Law’s National Legal Innovation Pitch Competition with it. 3L Jack Brandt, a Coast Guard lieutenant and one of National Jurist’s 2026 Law Students of the Year, built a Military Benefits Assistant at the LIT Lab that has served over 1,300 service members, veterans, and their families. He had no prior coding background when he started. Julia Rodgers, Suffolk Law Class of 2016, saw a problem with how prenuptial agreements were handled and built HelloPrenup, the first collaborative online prenup platform. She took it on Shark Tank. She won Suffolk’s Outstanding Graduate of the Last Decade award.
What each of them did was exceptional. But the path they carved is becoming wider. The barrier to building is lower now than it has ever been, and it is dropping fast. The next Julia Rodgers, the next Jack Brandt, the next Akila Narayanan might be sitting in a 1L contracts class right now and might build her first prototype on a weekday evening at a LITSA event.
Why Suffolk, why now?
Suffolk University Law School is the only law school in the country to hold four top-35 ranked legal skills specialties, in clinics, dispute resolution, legal writing, and trial advocacy, for ten consecutive years. Legal Writing is ranked #3 nationally. Clinical Programs are ranked #8. The school has climbed 40 spots in overall rankings over the past decade, the seventh-fastest rise of any law school in the country.
But rankings only tell part of the story. What makes Suffolk’s position distinctive right now is the infrastructure that was already in place when this moment arrived.
The Legal Innovation and Technology program started building before COVID forced courts online and legal education into remote formats. When that shift happened, Suffolk was ready. The concentration, the LLM, the clinic, the course offerings, the Document Assembly Line project that has helped tens of thousands of self-represented litigants, all of it was already operational or in development. That was not luck. It was institutional commitment over more than a decade.
Now a similar inflection point is here with generative AI, and Suffolk is positioned again. The ecosystem exists: faculty who understand the technology, students who are drawn to this work, clinical programs that deploy it in practice, and a culture that treats technological competence as a professional obligation rather than an elective interest.
A Suffolk alum who has led legal teams at three of the world’s largest technology and entertainment companies recently put it simply: innovation is a competency. It is a professional skill that belongs alongside legal research, client counseling, and courtroom advocacy.
That framing aligns with what the profession already requires. As Dean Andrew Perlman has emphasized, Comment 8 to Model Rule of Professional Conduct 1.1 establishes a duty to stay abreast of the benefits and risks associated with relevant technology. This is part of the baseline standard for competent representation. And the definition of “relevant technology” is expanding rapidly.
For prospective students thinking about where to study law: Suffolk offers a concentration, an LLM, a nationally ranked clinic, courses in AI regulatory frameworks and cybersecurity law, and a student body that is actively building legal technology and winning national competitions with it. Beyond the academics, there are great people here, a real culture of experimentation, and a prime location in downtown Boston with direct access to the courts, the State House, and a growing legal tech community. That combination does not exist at many other schools. It is worth paying attention to.
Making experimentation normal
The Innovation Challenge was designed to be low stakes. No prerequisites. No advance preparation. Show up with a laptop, pick a track, see what you can build. We wanted to remove every possible barrier between “I’m curious about this” and “I just tried it.”
What we found is that the distance between those two points is shorter than most people expect. Students who had never prompted an AI tool were producing structured legal analysis within their first few attempts. The gap between “I don’t know how to do this” and “I just built something that works” turned out to be about an hour.
That matters because experimentation is how competence develops. You have to try it yourself, see what works, recognize what falls short, and adjust. Every student who participated left with a better understanding of what these tools can and cannot do than they had when they walked in. That kind of learning does not come from a syllabus.
And some of them will keep going. There will be students who walk away from an event like this, or an internship, or a conversation, and realize what is possible. They will find a problem they care about, and they will build something to solve it. A tool that helps tenants understand their rights. A system that streamlines clinical intake. Something none of us have thought of yet. The capability is there, the problems are real, and law students are closer to the people these systems are supposed to serve than almost anyone else building technology right now.
What comes next
Suffolk has hosted legal technology events before, including hackathons in prior years. This year’s Innovation Challenge was a deliberate reset: designed to be accessible to everyone regardless of technical background, grounded in realistic legal scenarios, and focused on practical problem-solving rather than technical showmanship. The format worked.
We are planning to expand it next year into a larger, public-facing competition. If seven teams built this much in a single evening, we want to see what happens with more time, more participants, and more visibility. Suffolk should have its own signature legal tech event, and LITSA intends to build it. Any faculty, alumni, or practitioners who are interested in being involved, learning more about our plans, or connecting with our students should not hesitate to reach out.
But beyond the competition, what we are building is a culture. A culture where law students see themselves as people who can build things. Where experimentation is part of the curriculum and the extracurricular life of the school. Where students leave Suffolk with legal knowledge, practical skills, and portfolios of things they have actually built.
Suffolk has been leading in this space for over a decade. The students coming through this program are going to build things that matter: for their clients, for their communities, and for a legal system where the vast majority of people who need help are not getting it. The tools exist. The talent is here. The problems are not going to wait.
We are just getting started.