•  
• 
• Tweet  
•  
• Email
• Print

We’ve all seen the headlines about AI-boosted lawyers run amok. Since ChatGPT landed, phantom cases have cropped up in court filings around the country. Judges have responded, meting out sanctions, excoriating counsel, and—more recently—even issuing a flurry of new orders and rules that regulate how litigants can use new AI-based technologies.

But when it comes to lawyers’ use of AI, the solution is not bespoke new rules. Instead, as the ABA recently reminded us, it’s reliance on the decades-old regulatory architecture of attorney accountability. That time-tested architecture was up to the challenge, when, two decades ago, American lawyers started shipping legal work to professionals in India, and that delegation set off a short-lived ethics panic. It’s equally well-equipped to address the problem posed by lawyers’ reliance on AI, no amendments required.

If that’s all that was at stake, then we could say the frenzy to regulate use of AI has been a regrettable waste of courts’ time—and public resources. But the new spate of AI rules is worse than that. The new rules, it turns out, are affirmatively stunting innovative uses of AI that could help the millions of Americans without counsel. Worse still, they are distracting us from the more pressing problem: the need to reform the older, longstanding rules that restrict use of technology by everyone else, including courts themselves. Getting the rules right in that context is far more important to the future health of the system.

AI for lawyers. Everyone else? You’re on your own

It’s the dirty secret of American courts: Today, the majority of civil litigants are self-represented. Indeed, the best evidence suggests that, in something like three-quarters of the 20 million civil cases filed in American courts each year, at least one side lacks a lawyer. Most of these cases pit a lawyered-up institutional plaintiff (a landlord, a bank, a debt collector or the government) against an unrepresented individual. Facing highly consequential matters, from evictions to debt collections to family law matters, millions are condemned to navigate byzantine court processes—designed by lawyers, for lawyers—without formal assistance.

Of course, self-represented litigants aren’t entirely alone. Many are muddling through with the resources at their immediate disposal. Often, that means the internet or ChatGPT. Unfortunately, though, both are chock-full of unreliable legal information. As the National Center for State Courts recently put it, in the age of AI, the American legal system is increasingly awash in a “sea of junk.”

All is not lost just yet. As with so much else in AI, the situation is in flux. Generative AI tools are getting better, fast. When we squint, we can glimpse a not-too-distant future where AI tools offer real, valuable assistance to self-represented litigants. Even improved tools, however, will run into barriers.

Two barriers stunting generative AI’s capacity to help self-represented litigants

One barrier is longstanding rules in every state, dubbed unauthorized practice of law rules (or UPL rules for short), that say only lawyers can practice law—and then define “practice of law” capaciously. Those rules apply even to nonhumans and so prevent tech providers—the Legal Zooms of the world—from offering comprehensive assistance to people who need it. UPL rules are already stunting tech tools’ ability to help self-represented litigants. And UPL rules’ limiting effect will only intensify as the capabilities of tech tools grow.

Then, there are those new rules that courts and judges, in their AI fever, are hastily promulgating. Consider a recent order from a federal court in North Carolina. It prohibits the use of AI in research for the preparation of a court filing “with the exception of such artificial intelligence embedded in the standard on-line legal research sources Westlaw, Lexis, FastCase and Bloomberg.”

Can you guess how many unrepresented litigants have access to these pricey commercial databases? “Not many” would be an understatement. Essentially, this order gives lawyers the green light to use generative AI while tying the hands of those without counsel.

Some rules are less heavy-handed and merely require the litigant to disclose any AI use. But even these can have a chilling effect, particularly when it comes to litigants without counsel. Do self-represented litigants have to disclose if they use a search engine with generative AI capabilities? Will the average person even know? What if someone simply used a generative AI tool to parse the thicket of legalese that dots court websites? Must that be disclosed? The answer to these questions is unclear, which highlights the burdensome and restrictive nature of these knee-jerk policies.

‘Courthouse AI’ as the new frontier of access to justice

What to do? One can readily imagine lawmakers and rulemakers responding to hallucination and sea-of-junk concerns by doubling down on UPL provisions and prohibiting OpenAI and others from doing law. Lawyers focused on their bottom lines might applaud that development.

But there may be a better option—and it’s already underway. Courts are beginning to incorporate AI into their own operations, and they are positioning themselves as an authoritative source of legal information and self-help resources. Harnessing generative AI, courts can make it so their websites, portals and conveniently located kiosks furnish reliable, actionable and individually tailored information to self-represented litigants. Newly digitized courts may be the only institutions positioned to serve as a life raft to keep self-represented litigants afloat in the sea of junk.

The catch? The same UPL rules that hamstring the Legal Zooms of the world also bar courts and courthouse personnel from giving self-represented litigants reliable, actionable and tailored advice, under threat of criminal penalty. This restriction—what we call “courthouse UPL”—presents a considerable obstacle to positioning our nation’s courts as trusted, authoritative sources of legal guidance for unrepresented parties. It also limits the digital assistance that courts can provide.

Solving this issue is more challenging, of course, than issuing orders narrowly targeting lawyer brief-writing. We need to update the outmoded guidelines that states have created declaring what courts can and can’t do. Many of those guidelines speak to an earlier, analog era—and even the more recent ones address the static websites of yesteryear, not the dynamic, interactive tools that generative AI makes possible.

Lacking technical capacity of their own, courts also need to develop AI R&D pipelines, whether via smart procurement or by working with universities and a growing “public interest technology” movement, to learn what works and develop court-hosted tools that are trustworthy, flexible and responsive to litigant needs. Given AI’s promise for the millions of Americans consigned to navigate courts without help, we must confront questions about the court’s role and “courthouse UPL” head-on. Courts are neither neutral nor impartial if they choose to restrict rather than facilitate litigant access to AI-based assistance.

Time will tell what a new, digitized civil justice system will look like. What’s clear, however, is that when it comes to lawyer use of AI, the existing attorney regulatory architecture is already adequate. No further guidance is required. For self-represented litigants and the courts that are laboring to serve them, the story is different. For them, generative AI holds real promise if we’re wise enough to let it.

David Freeman Engstrom is the LSVF professor of law at Stanford Law School. Nora Freeman Engstrom is the Ernest W. McFarland professor of law at Stanford Law School. They co-direct Stanford’s Deborah L. Rhode Center on the Legal Profession.

ABAJournal.com is accepting queries for original, thoughtful, nonpromotional articles and commentary by unpaid contributors to run in the Your Voice section. Details and submission guidelines are posted at “Your Submissions, Your Voice.”

This column reflects the opinions of the author and not necessarily the views of the ABA Journal—or the American Bar Association.

Give us feedback, share a story tip or update, or report an error.