A new book edited by Stanford Law School’s David Freeman Engstrom, the LSVF Professor in Law and co-director of the Deborah L. Rhode Center on the Legal Profession, takes a deep dive into technological developments in the legal system, from virtual legal proceedings to AI-fueled litigation tools. In Legal Tech and the Future of Civil Justice, Engstrom and his 28 co-contributors, including six SLS faculty members, dissect the legal and policy implications of the technologies that are poised to remake the civil justice system.
Avoiding ultra-futuristic prognostications around “robo-judges” or “robo-lawyers,” the book instead focuses on what is already here or coming around the corner. As Engstrom notes in his introduction, the book seeks to enrich, and also reorient, the debate about legal technology’s implications for the civil justice system “by offering a grounded, concrete, empirically minded discussion of the current state of legal tech and what, actually, lies ahead.”
In this Q&A, Engstrom discusses some of the arguments and assertions in Legal Tech and the Future of Civil Justice, which grew out of a 2021 SLS conference of the same title. The book, published by Cambridge University Press, is available for free download via Open Access on Cambridge Core.
In addition to Engstrom, the SLS contributors to the book are: Nora Freeman Engstrom, the Ernest W. McFarland Professor of Law and co-director of the Rhode Center; Margaret Hagan, executive director of the Stanford Legal Design Lab; Julian Nyarko, associate professor of law; Norm Spaulding, the Nelson Bowman Sweitzer and Marie B. Sweitzer Professor of Law; Todd Venook, a fellow at the Rhode Center; and Diego Zambrano, associate professor of law.
Can you illuminate some reasons why the development of legal technology has seemingly accelerated so quickly in recent years?
In the legal system, as in so many other aspects of our world, the COVID-19 pandemic was a powerful accelerant of certain trends already in motion. Pre-COVID, Zoom proceedings were a rarity for most lawyers, but also plainly making their way into the system—via, say, remote depositions. As a result of the pandemic, online proceedings are exploding, evolving, and finding their place in a variety of courts. An important example is online dispute resolution (ODR) platforms. These are not live legal proceedings, but online platforms where disputants can gather, often asynchronously, and bargain their way to settlement. ODR is now implemented or piloted in at least 200 state and local court jurisdictions, with more to come.
Another reason is continuing technological advances, particularly Natural Language Processing (NLP), the branch of machine learning that performs text analytics and so holds the most promise in a legal system that trades in words. NLP has seen some drastic improvements in recent decades—ChatGPT is only one recent and highly visible development—and Julian Nyarko’s chapter, co-authored with Washington University’s Jens Frankenreiter, expertly explains how NLP-powered tools will steadily improve in their ability to perform higher-order legal tasks. Indeed, NLP is gaining rapidly in its ability to extract and process information from the sea of unstructured text that lawyers routinely encounter, whether judicial opinions, statutes, legal briefs, written agreements, or witness testimony. However, as Nyarko and his co-author note, there are also, for the moment, real limits on what NLP can do. Humans must still provide the “legal ontologies”—basically, knowledge representations of doctrinal structures and tests—before higher-order legal tasks such as case outcome prediction can be automated. Legal tech tools still very much depend on lawyers and their unique skill sets and capacity to understand legal reasoning. Still, it seems clear that, with further advances, NLP-powered tools can be expected to outperform humans in a growing set of legal tasks, particularly where large amounts of pre-labeled data are available or can be generated to train an algorithm.
These developments raise fascinating—and often worrisome—questions that go well beyond the role of technology, challenging the definition of what it means to be a lawyer.
Online legal proceedings, including virtual trials, would seem to be one of the most notable pandemic-era changes to the legal system. Are online proceedings here to stay, and what issues do they raise?
Our book offers sharply contrasting views on the move to remote proceedings, which are indeed here to stay. Several contributors to the book, led by James Greiner and Renee Danser from Harvard Law, surveyed the existing social science literature on virtual interaction in both law and non-law contexts. Among other things, they conclude that there is little reason for concern that the “Zooming” of litigation degrades the already-low truth-detection capacities of judges or juries. By contrast, a chapter by Victor Quintanilla from Indiana Law School sounds a more skeptical note. Building off his work in law and psychology and the social construction of pro se litigants, Quintanilla offers some initial findings from an ongoing study of Zoom courts in Indiana and shows how virtual hearings can exacerbate distributive concerns because of the lack of access to, or facility with, technology among already-marginalized groups.
It is easy to imagine from there a concrete catalog of concerns raised by the online migration of legal proceedings: rural communities without access to high-speed internet, large numbers of self-represented litigants with low technical literacy, and use of outmoded court technologies.
A major focus of your scholarship is on the millions of Americans who, for financial and other reasons, cannot get their day in court or otherwise leverage the legal system. Can technology help them?
That’s an important, complex question and a focus of the book. The pandemic shined a harsh light on what was already a growing recognition that American courts, particularly state ones, are in the throes of an access-to-justice crisis. In some three-quarters of the 20 millions civil cases filed in state courts each year, at least one side lacks a lawyer. That means millions of people who are facing debt collection, eviction, and difficult family law matters must try to represent themselves, often against well-resourced, represented adversaries. And these are only the visible litigants who show up on court dockets. Beneath them are tens of millions more who are locked out of justice entirely, and are often not even aware that they have legal problems in the first place.
More and more states are taking stock of this depressing landscape and are considering deregulating legal services by relaxing the usual rules that say only lawyers can practice law or own law firms in order to bring new providers into the system, including the software variety. Online legal help and document assembly tools that help pro se litigants create and file legal documents can be helpful. These tools can be important not just for self-represented litigants, but also for the smooth operation of our courts.
Websites like LegalZoom and Rocket Lawyer, as well as a growing array of smaller self-help websites, aim to make the “legal journeys” of people who are unable to retain counsel less burdensome and more effective. But as Margaret Hagan notes in her chapter on “The Supply and Demand of Legal Help on the Internet,” the mere existence of legal aid websites will not by itself meaningfully narrow the justice gap. People must also be able to find and effectively use these sites to build legal capacity, Hagan writes. That can be hard when the internet is dominated by commercial sites that often engage in questionable practices, such as baiting users with incomplete information and then charging for more. Here, far more research is needed to understand both the supply and demand side of legal services and also possible policy interventions, including ways we might induce search engine companies to ensure that online searches return trustworthy and actionable legal information.
Another significant problem is the decentralized nature of courts, which contributes to a checkerboard of different technology infrastructures and data standards across thousands of state and local jurisdictions. The resulting variation is a problem because it defeats the scale that technology providers who wish to assist self-represented litigants need in order to invest in the creation of robust, user-friendly tools. That, by the way, is the theory at the heart of the Filing Fairness Project, a Rhode Center initiative that is working with six states to standardize their technical systems and reduce barriers for self-represented litigants to make it easier for technology providers to serve them.
Various chapters of your book would suggest that technology is not going to be a magic bullet with regard to closing the justice gap. Is that correct?
Yes. In our chapter, “Legal Tech and the Litigation Playing Field,” Nora Freeman Engstrom and I challenge the view that technology will be a great equalizer within the civil justice system. We agree that some technology tools may pay dividends for both plaintiffs and defendants, promoting transparency, efficiency, access, and equity. Legal tech tools can be a force-multiplier, allowing smaller firms to do battle with larger ones. If one of the causes of the access to justice crisis is a shrinking PeopleLaw—the segment of the legal services industry that serves individuals and small business, as against BigLaw’s corporate focus—then legal tech’s efficiencies could powerfully increase access by lowering the cost of providing legal services, bringing them more within reach of middle and moderate income Americans.
But other, arguably more powerful, tools are also coming on the scene. And many of the most potent ones are, and will remain, unevenly distributed among litigants. This is because large and often corporate repeat-players within the system will alone have access to high-quality data on case outcomes in a system where most cases end in unfiled, secret settlements. Far from leveling the playing field, the innovation ecosystem may, at least over the near- to medium-term, give yet more advantages to the “haves.” Powerful repeat players, leveraging their privileged access to confidential claim-settlement data and their ability to mine and deploy that data, will propel themselves yet further ahead.
Can you share an example of a specific legal technology that purports to be an equalizer, but may do the opposite?
I mentioned online dispute resolution platforms earlier, and the book contains a bracing exchange between JJ Prescott, a professor at the University of Michigan Law School and a pioneer of ODR platforms, and my Stanford colleague Norm Spaulding, who is especially good at thinking about the historical arc of law, technology, and the legal profession. In his chapter, “Online Dispute Resolution and the End of Adversarial Justice?,” Professor Spaudling takes issue with proponents of ODR systems who say that their innovations will expand access to justice for so-called “simple cases.” The argument that ODR helps settle “simple cases,” Spaulding notes, assumes that the dollar value of a dispute is a reliable proxy for its complexity and its significance to the parties. But this assumption may be wrong. The huge mass of debt collection actions and evictions that dominate state court dockets might be low-dollar compared to certain other cases within the civil justice system, but they can be hugely consequential for the parties. Many of these cases are brought by well-represented private or public institutions, whereas the defendants are overwhelmingly unrepresented individuals.
What ODR offers in these settings is not access to justice for ordinary people, Spaulding argues, but rather a powerful compliance technology for private creditors and the state. And, while resolution of some disputes may be more expeditious, parties will often lack meaningful information about their rights and defenses. As with so much else in technology policy, the devil will be in the details. Courts creating ODR platforms face very significant design and oversight challenges—and, without significant technical capacity of their own, may not be well-equipped to meet them.
Much of what we’ve discussed concerns technologies that serve self-represented litigants. What types of technologies are being used by lawyers to serve clients?
The main place where AI and machine learning are already in use within the civil justice system is Technology Assisted Review (TAR), a powerful technology that assists lawyers with document review. In a nutshell, lawyers label a small subset of millions of documents for relevance or privilege, then use them to train a machine learning algorithm that can label the rest. The tools have the potential to revolutionize a litigation system that has been dogged by concerns about high costs, which can shape settlement as much as case merits. Diego Zambrano and a pair of JD/PhD students in computer science here at Stanford contributed a chapter that looks at how these new tools might contribute to new, technological modes of gamesmanship in discovery. Professor Zambrano raises important questions about TAR, how AI might exacerbate gaming in a system where technological expertise is often unevenly distributed across the “v,” and how we might adapt analog procedures to address it.
Can you talk a little about your own tech-focused project to bridge the justice gap?
The Filing Fairness Project, which I mentioned earlier, is an initiative we have been working on at the Rhode Center, along with the Stanford Legal Design Lab. This is a multi-jurisdictional effort to simplify court filing processes and improve access to and the administration of justice by leveraging readily available technology.
Plain-language interview systems to gather information and generate complex forms already exist for tax filings, mortgage applications, and benefits administration. But court filings have lagged behind these applications for a host of reasons, including institutional history and inertia. Because it is cost-prohibitive to create solutions jurisdiction by jurisdiction, we are working with several state court systems to develop multistate online solutions.
The goal is to provide accessible e-filing options for litigants and, by extension, help improve the accuracy and relevance of court filings and increase access to justice. The initial focus is on name change petitions, fee waiver requests, and eviction answers—filings where self-representation and related challenges are particularly pronounced.
David Freeman Engstrom is a far-ranging scholar of the design and implementation of litigation and regulatory regimes whose expertise runs to civil procedure, administrative law, constitutional law, law and technology, and empirical legal studies. He co-directs the Deborah L. Rhode Center on the Legal Profession, the premier academic center working to shape the future of legal services and access to the legal system. Professor Engstrom’s work focuses on access to justice in the millions of low-dollar but highly consequential cases, including debt collection, eviction, foreclosure, and child support actions, that shape the lives of Americans each year. He currently serves as the Reporter for the American Law Institute’s Principles of the Law, High-Volume Civil Adjudication, which will offer courts guidance on the challenges these cases raise.