How AI Is Changing Self-Represented Litigants, and What It Means for the Courts

A self-represented litigant just prepared a motion approaching 300 pages with the help of ChatGPT.

That may sound extraordinary today, but I don't think it will for long.

I was talking recently to someone who has been forced to represent herself in a family law matter....and not because she wants to.

The father of her child has significantly more financial resources than she does. Every time she asserts her rights, the legal fees pile up. Eventually, the economics of litigation become part of the problem: she simply cannot afford to keep up.

So, she turned to AI.

She's an incredibly smart person to begin with. She has spent months learning the process, researching the law, organizing years of evidence and trying to understand how to properly put her case before the court.

ChatGPT didn't create the facts for her, and it didn't suddenly turn her into a lawyer, but it gave her something incredibly important: the ability to work through an enormous amount of information that she otherwise would have needed significant financial resources to properly organize and understand.

And now she has a motion approaching 300 pages.

The more we talked, the more I realized that the interesting part isn't the 300 pages. It's what those 300 pages represent.

She isn't the exception. She's the preview.

AI Is Changing What Self-Represented Litigants Can Do

For years, there has been an enormous capability gap between someone represented by experienced counsel and someone standing in court alone.

Money didn't simply buy someone to appear in court. It bought hours of legal research, document review, drafting, organization, issue spotting and strategy. It bought someone who knew what to look for and who could take years of emails, text messages, financial records, previous orders, affidavits and correspondence and turn that mountain of information into something a court could actually work with.

A self-represented litigant often had to figure out how to do all of that alone.

AI changes part of that equation.

Someone can now use AI to help explain unfamiliar terminology, organize documents, create chronologies, compare statements, identify questions to research and make sense of an overwhelming record.

That doesn't eliminate the enormous advantage that legal representation provides.

Someone with substantial financial resources can still retain experienced counsel, pay for experts, respond to every application and absorb the cost of lengthy proceedings. AI doesn't suddenly make the playing field financially equal.

But it does give the person on the other side capabilities they didn't have before, and I think that's an incredibly important development for access to justice.

The person who previously ran out of money and effectively ran out of options may now be able to understand what is happening, organize their evidence, research the issues and participate more meaningfully in their own case.

Now, let's be clear, AI ISN'T a lawyer.

But access to information, organization and understanding shouldn't belong exclusively to people who can afford unlimited legal fees.

That matters.

Now Multiply Her by Millions

This is where the story gets bigger than one self-represented litigant.

Go back to that motion. One person, one part of one case, nearly 300 pages.

Now imagine what happens as these tools become ordinary.

People who previously couldn't afford to have five years of correspondence reviewed and organized may be able to do a meaningful portion of that work themselves. People who don't understand the terminology in a court document can ask what it means. Someone facing thousands of pages of disclosure can use AI to help make sense of it.

Lawyers are gaining the same capabilities at an entirely different scale. They can research faster, analyze larger records, compare more documents and produce sophisticated work in significantly less time.

That's the part of the AI talk I think we are alll significantly underestimating.

AI doesn't just save work. It creates capacity.

And when you give millions of people more capacity, they don't necessarily produce less, they can produce more.

More developed arguments, more evidence organized into usable form, more issues identified, more documents analyzed and more material capable of making its way into a proceeding.

All of that has to go somewhere.

It goes into the courts.

If technology dramatically increases the capacity of the people entering the justice system without increasing the capacity of the institutions receiving all of that information, we haven't eliminated the bottleneck.

We've moved it.

The Courts Need AI Too

A judge cannot solve an information-volume problem simply by reading faster.

If lawyers can research faster, analyze more information and produce more work, and self-represented litigants can do the same, courts will increasingly be asked to process the output of all that new capacity.

A judge may already be dealing with hundreds or thousands of pages containing competing allegations, affidavits, exhibits, previous orders, correspondence and legal authorities.

Now give everyone involved better tools to produce and analyze even more....the obvious conclusion isn't that courts somehow remain untouched by AI.

They need it too.

We're already beginning to see that direction emerge.

Clio's acquisition of Learned Hand, a company building AI technology for judges and courts, is one example of legal AI moving into the judiciary.

The acquisition itself isn't really what interests me, but what it signals does.

AI could eventually help courts navigate enormous records, construct chronologies, connect allegations with evidence, identify inconsistencies, surface previous orders and locate important information buried hundreds or thousands of pages into a case.

That could be transformative.

It could also become necessary, because if AI gives everyone outside the courtroom the ability to produce and analyze more information, someone has to give the court the ability to deal with it.

But that's where we run into the part of this that I find most interesting.

What Happens When AI Starts Talking to AI?

Imagine the same family law case a few years from now.

The self-represented litigant uses a general AI tool like ChatGPT to help her work through her documents, evidence and history, organize the record and understand the legal issues she needs to research.

On the other side, the lawyer and law firm may be using specialized legal AI  tools built specifically for legal research, document analysis, discovery, case strategy and working across large matter files.

Those are very different systems.

The lawyer's AI may have access to authoritative legal databases, the firm's own knowledge and the evidence contained within the client's matter. The self-represented litigant's AI may be working primarily from the information and documents she provides.

Both sides are using AI to analyze evidence, build timelines, identify inconsistencies and make sense of increasingly large amounts of information.

Then all of that arrives at the court, where judicial AI may eventually help a judge navigate the record, connect allegations with evidence, compare competing accounts and understand thousands of pages.

Now we have different AI systems, built for different users and working from different sources, effectively feeding information into one another through the justice system.

AI starts talking to AI.

And that's where this gets really interesting because the biggest issue here isn't whether all of these systems are equally sophisticated, it's that none of them has independent access to the objective truth of what happened.

The self-represented litigant's AI knows the factual record it has been given. The law firm's AI knows the factual record it has been given. And the court's AI ultimately knows the record that makes its way before the court.

Every system can become extraordinarily good at analyzing its version of the record without any one of them possessing the whole truth.

AI Doesn't Know What Actually Happened

This is the distinction I think we're going to have to become much more sophisticated about.

We spend a lot of time talking about whether AI can accurately research the law.

That's important, but the law has sources.

A legal proposition can be traced back to legislation, a regulation, a court rule or a judicial decision. If an AI says a case stands for something, we should be able to go back to that case and determine whether it actually does.

Facts don't work the same way.

ChatGPT doesn't have some independent database containing the objective truth about what happened between the two people in the case I'm talking about.

It has what she gives it.

Her emails, his emails, text messages, financial records, previous orders, affidavits, disclosure, photographs and whatever other evidence she has become the factual universe the AI is being asked to analyze.

Inside that universe, AI can do extraordinary things.

It can build a chronology from years of documents, compare statements made at different times, connect an allegation with an exhibit and identify something in an affidavit that appears inconsistent with a text message sent months earlier.

But it can only work with what it has.

It doesn't know about the email that was deleted, the income that wasn't disclosed, the conversation that was never recorded, the document nobody produced or the evidence one side doesn't know exists.

And it certainly doesn't know that something is true simply because someone said it happened or swore it in an affidavit.

Now imagine that problem on both sides.

Her AI may have one factual universe. His lawyer's AI may have another. Both systems could function exactly as intended. Neither one needs to hallucinate. Neither one needs to make a technical mistake.

And they could still help produce two completely different versions of what happened.

Because they were given two different records.

The Court's AI Has the Same Problem

Putting AI inside the court doesn't magically solve this.

It may give the court an extraordinary ability to analyze what has been filed.

Imagine a judge being able to click on a factual statement and immediately see the affidavit paragraph, original email, text message, bank record or exhibit said to support it.

Imagine the system also surfacing evidence elsewhere in the record that contradicts that statement.

I think that's where this technology could become incredibly powerful, but even the best judicial AI can only analyze the record that exists before it.

If an AI-generated summary says that something happened on March 14, the important question isn't simply whether the AI accurately summarized the documents.

It's why the AI believes it happened on March 14.

Did that date come from a contemporaneous text message, a bank transaction, an email, an affidavit sworn three years later, an allegation in a pleading or someone else's summary?

Those aren't equivalent sources.

And what happens if the one document that completely changes the context was never put before the court?

The AI doesn't know what it doesn't know. Neither does the judge.

That's not actually a new problem in litigation. Courts have always had to make decisions based on the evidence put before them, AI just gives us the ability to process that evidence at a scale we've never had before.

And that makes knowing where the information came from even more important.

A Coherent Story Is Not Necessarily a True Story

Generative AI is remarkably good at structure.

Give it thousands of pages and it can turn scattered information into a coherent narrative.

That's enormously useful, but it's also exactly why we need to understand its limits.

Litigation isn't simply a competition between two beautifully organized stories. The underlying facts matter, evidence matters, context matters and credibility matters. Sometimes two people genuinely remember the same event differently. Sometimes a document contradicts someone's recollection. Sometimes evidence is missing.

And sometimes someone lies.

AI may become extraordinarily good at identifying contradictions inside the information it receives, but identifying a contradiction isn't the same thing as determining credibility.

Organizing a story isn't the same thing as proving it.

And making something coherent doesn't make it true.

AI Isn't Removing the Humans. It's Giving Them More to Do.

This is why I don't think the most interesting question is whether AI will replace lawyers or judges.

I think we're asking the wrong question entirely. AI is increasing what everyone in the justice system can do.

A self-represented litigant can do more, a lawyer can do more, a law firm can do more, and eventually, a court will be able to do more.

And when everyone has more capacity, the result may not be less legal work.

It may be significantly more.

A person who previously couldn't afford to pursue an issue may now be able to. A lawyer who previously didn't have the time to analyze 10,000 documents may now be able to. An argument that would have taken days to research may take hours. Evidence that might have remained buried may be found. More issues may be identified. More sophisticated material may reach the court.

AI may make individual tasks dramatically more efficient while simultaneously increasing the overall volume and complexity of the work being done.

That's not replacement, it's full on expansion, and the more AI becomes involved, the more important another human function becomes.

Accountability.

Someone Still Has to Be Accountable

This is the part we can't automate away.

AI can help a self-represented litigant understand her case, help a lawyer analyze it, and help a judge navigate it, but someone still has to stand behind what happens next.

Lawyers have professional and ethical obligations. They decide what arguments to advance, advise clients about risk, determine how evidence should be presented and are accountable for the work they put before a court.

Judges carry an even more fundamental form of accountability.

A court cannot ultimately say:

The AI decided.

Someone has to determine what evidence should be accepted, assess credibility, apply the law, exercise judicial discretion and take responsibility for the decision.

The more powerful these systems become, the more important that line of accountability becomes.

AI can make information easier to find, easier to organize and easier to understand. It can show us relationships in the evidence that a human might otherwise miss. It may eventually allow a judge to understand an enormous case faster and more completely than would have been possible before.

But it doesn't become responsible for the consequence.

People do.

And That Brings Me Back to the 300 Pages

The woman I was talking to used AI because she couldn't afford to keep up.

For her, this isn't an abstract discussion about the future of legal technology.

It's access.

AI helped her work through years of information, understand what she was dealing with and organize a case that is now approaching 300 pages. Without it, the financial imbalance between the parties doesn't just affect who has the better lawyer.

It affects who has the capacity to participate, and that's why I'm optimistic about this technology.

Not because I think it eliminates lawyers, and definitely not because I think we should hand decisions to machines, but because knowledge, organization and the ability to understand your own legal problem shouldn't be available only to the person who can afford the most hours.

Soon, I don't think her story will be unusual.

Neither will the lawyer on the other side using AI to analyze what she files, and eventually, neither will the court using AI to help make sense of what both sides put before it.

That's where this is going.

AI is going to dramatically increase what people throughout the justice system are capable of doing.

It may also dramatically increase the amount of work the system produces, and somewhere in that increasingly AI-assisted chain, a human still has to be responsible for what happens.

In the end, AI can organize, analyze and understand....

But the justice system still has to determine what actually happened, and someone has to be accountable for that decision.

Frequently Asked Questions About AI and the Courts

Can self-represented litigants use AI to help with court cases?

AI tools can help self-represented litigants understand terminology, organize information, summarize documents, create timelines and identify issues they may need to research.

That could have significant implications for access to justice because some of the capabilities that historically required substantial amounts of professional time are becoming available to people who may not be able to afford continuous legal representation.

AI doesn't turn someone into a lawyer, however, and AI-generated information can be inaccurate. Court procedures and laws also vary by jurisdiction, so its output still needs to be treated carefully.

Could AI improve access to justice?

Potentially, yes.

Cost is one of the barriers that can prevent people from participating meaningfully in the legal system. AI can reduce some of the time and specialized knowledge required to understand documents, organize evidence, research issues and prepare information.

It doesn't eliminate the value of legal representation or solve the financial imbalance between litigants, but it can give people who cannot afford extensive legal assistance capabilities they previously didn't have.

Where does AI get the facts of a person's case?

When AI is being used to analyze an individual legal matter, much of the factual information comes from the person using it and the materials provided to the system.

Those materials may include affidavits, emails, text messages, financial records, court orders, exhibits, disclosure and descriptions of events.

AI can analyze that information, but it cannot independently establish that every factual assertion is true or know whether important information is missing.

Can AI tell whether someone is lying in an affidavit?

AI may be able to identify inconsistencies between an affidavit and other available evidence, but identifying a contradiction is different from determining credibility or truth.

The system may also have no way of knowing that relevant evidence was omitted, destroyed, never recorded or never provided.

What happens when both sides of a legal case use AI?

Each side's AI may be working from a different collection of documents, evidence and factual assertions.

Both systems could accurately analyze the information they receive and still help produce very different accounts of what happened because the underlying factual records are different.

That makes the ability to trace AI-generated conclusions back to the underlying evidence increasingly important.

How could AI help judges?

AI could help with information-heavy judicial work such as navigating large court records, constructing timelines, identifying disputed issues, connecting allegations with evidence, locating inconsistencies and researching legal authorities.

That doesn't mean the technology becomes the decision-maker. Its value may be in helping judges process increasingly large and complex records while leaving judgment and accountability with the court.

Why could AI create more work for the courts?

Because efficiency and volume aren't the same thing.

AI may make it dramatically faster for an individual person or lawyer to research an issue, analyze evidence or prepare material. But if millions of people gain those capabilities, the total amount of legal work being produced can increase.

More people may be able to pursue issues, more evidence may be analyzed, more arguments may be developed and more sophisticated material may reach the courts.

AI can therefore make individual tasks more efficient while increasing the overall volume of information the justice system has to process.

Will AI replace lawyers?

AI will automate and accelerate a significant amount of work lawyers currently do, including research, drafting, document review, evidence analysis and case organization.

It will not eliminate the lawyer.

Someone still has to stand behind the advice, determine strategy, decide what arguments should be advanced, ensure evidence is properly presented, meet professional and ethical obligations and be accountable to the client and the court.

AI may actually contribute to more legal work rather than less by allowing people to identify more issues, analyze more evidence and participate more actively in legal matters.

The role of the lawyer will change as AI becomes more capable.

The accountability doesn't disappear.

Will AI replace judges?

AI may eventually become extraordinarily capable at reading court records, constructing timelines, comparing evidence, identifying inconsistencies and helping judges navigate enormous amounts of information.

But a court can't simply say, "the AI decided."

Someone still has to assess the evidence, determine credibility, apply the law, exercise judicial discretion and take responsibility for the decision.

AI may also give judges more to deal with, not less. If lawyers and self-represented litigants can produce more sophisticated arguments and analyze much larger volumes of evidence, courts may receive substantially more information.

AI can increase judicial capacity.

It can also increase the amount of work reaching the judiciary.

The judge remains because someone has to be accountable for the decision.

____________________________________________________

About the Author

Cara Echino is the founder of Lawggle, a legal technology platform focused on modern legal discovery and the future of how people find legal help online. With decades of experience in the legal industry, she writes about legal technology, AI, access to justice, lawyer visibility, and the changing relationship between people, lawyers and the legal system.