In today’s column, I examine the intriguing and beguiling situation whereby AI hallucinations are utilized and relied upon by attorneys on both sides of a court case. This is rare, but will arise on an increasing basis due to attorneys continuing to improperly make use of AI for generating aspects of their legal briefs and failing to double-check their work.
The question at hand is what should happen if both sides of a case commit the same type of faulty professional efforts involving AI hallucinations at the same time?
Let’s talk about it.
This analysis of AI breakthroughs is part of my ongoing Forbes column coverage on the latest in AI, including identifying and explaining various impactful AI complexities (see the link here).
Use Of AI Hallucinations In Court Cases
As readers well know, I’ve been extensively covering and analyzing a myriad of facets regarding the intersection of AI and the law for many years. You can find my writings not only in my Forbes column but also as posted in Bloomberg Law, ABA Law Journal, The National Jurist, The Global Legal Post, Lawyer Monthly, The Legal Technologist, MIT Computational Law Journal, and so on.
I have extensively examined the repercussions for attorneys who have used AI-hallucinated elements in their legal briefs; see my coverage at the link here and the link here. The situation usually is that an attorney opts to include fake legal citations or fictitious quotations about legal cases, having generated those via the sloppy use of generative AI and large language models (LLMs).
I’ve frequently noted that until the courts impose stiffer penalties, or until lawyers wake up and use suitable double-checking procedures and tools, this phenomenon is going to continue. There is a constant drumbeat of lawyers contending they had no idea that AI could produce bad references or bogus quotes. Though this might have seemed plausible when generative AI first hit the scene, the claims of innocence now seem quite hollow. Most savvy attorneys raise their eyebrows when a caught red-handed lawyer tries to plead ignorance of these matters. It does not compute anymore.
Background On The Happenings
Daily bulletins in the legal community keep highlighting lawyers who have used AI to prepare their court filings and ended up with fake legal citations and false quotations in their documents. These errors in legal briefings can potentially occur due to AI hallucinations. An AI hallucination is when generative AI or large language models (LLMs) such as ChatGPT, Claude, Grok, Gemini, CoPilot, Llama, and other AIs veer into generating fictitious confabulations. For my in-depth coverage of AI hallucinations, see the link here and the link here.
I had long ago predicted that attorneys using AI might get careless in their legal efforts and allow the AI to produce bogus content and not double-check by hand what the AI has generated for them (see my prediction in 2023, at the link here).
The problem for attorneys is that when they submit formal court filings, they are supposed to be responsible for the contents of the filings; thus, if AI has slipped in faked citations or false quotations, the lawyer is likely to be held accountable since they didn’t catch the erroneous content. I say “likely” accountable because judges and courts have been quite lenient so far, overall, and allowed excuses such as “the computer did it”. Attorneys often incur nothing more than a minor hand slap or mild rebuke, asserting that AI is new to the legal beagles and they were caught unawares. Sometimes, the reprimands are accompanied by a modest financial sanction or penalty, which is gradually ratcheting up as these instances continue to climb.
Judges and courts are beginning to lessen their patience and sense of charity in giving lawyers the benefit of the doubt. For the time being, the matter is still generally being treated with kid gloves. The expectation is that lawyers are facing a learning curve, and a bit more time is required before they will be able to properly handle the use of AI. In my view, more than enough time has already passed.
Stop coddling legal pros that are supposed to remain on top of the latest technological aspects when performing their legal services (see my coverage of ABA Formal Opinion 512, at the link here, which was promulgated on July 29, 2024, nearly two years ago!).
Twists And Turns Aplenty
I recently examined the somewhat rare situation of attorneys that fail to detect that the opposing side has made use of AI hallucinations in their legal filings in a court case; see the link here. The question there is whether the attorney that doesn’t catch the mistakes of their opposition ought to receive any sanctions or penalties. There are tradeoffs involved, which I analyzed at length in my posting.
There is another new twist that is garnering attention, namely the exasperating circumstance of both sides of a case relying on AI hallucinations at the same time. You might be shocked to think that this could happen. It’s one thing for an attorney on one side to make such an error, but having both simultaneously do this is astonishing and dismaying.
On the other hand, given that attorneys seem to be ignoring the blaring headlines or otherwise opting to fall asleep at the wheel when using AI, the chances of both sides committing the same type of error are perhaps higher than we might want to admit. Sliding into quicksand is heightened when both parties opt to stand ill-prepared at the same edge.
The Variations At Play
Consider these three major variations that can arise:
- (1) Court detection: Attorneys on one side use AI hallucinations, attorneys on the opposing side use AI hallucinations, and the court figures this out.
- (2) One-side attorney detection: Attorneys on both sides use AI hallucinations, one side tells the court the other side has done so, not realizing they also did so.
- (3) Both sides detection: Attorneys on both sides use AI hallucinations, both sides tell the court that the other side has done so, each respective side might or might not fess up that they did the same.
The first instance involves the court figuring out that both sides have relied on AI hallucinations. This is a somewhat more straightforward circumstance. Both sides can be held accountable on a relatively equal basis for their respective conduct. This includes that neither side detected on their own that they made use of AI hallucinations, which is bad, and worse still, that neither side detected that the opposing side made use of AI hallucinations.
It seems especially egregious that neither side detected the use of AI hallucinations by the opposing side. You would think that the intrinsic nature of our adversarial system would ensure that at least one side would detect that the other side is using AI hallucinations. Both sides are supposed to be scrutinizing the filings of the opposing side. Failing to do so is quite a failing of duty to the adversarial approach and to the serving of justice and the courts.
Pox on both heads.
The One-Sided Reveal
If the circumstance entails one side detecting that they themselves used AI hallucinations and then telling the court accordingly or finding that the opposing side has done so and reporting this to the court, it is seemingly a little bit better than the court having to do the detective work entirely on its own. That being said, only in the situation of both admitting to AI hallucinations and pointing out that the opposing side also made use of AI hallucinations would there seem to be a chance for merit of some kind.
The gist is that if one side only reports itself or the opposing side, they have still missed the bigger picture. They did not detect that both sides faltered.
I’m not sure whether much credit should go toward only figuring out that one side messed up. There are arguments either way. You could claim that at least the one side found something, while the other side was apparently completely in the dark. Either way you stack it, this seems like the classic notion of moving around the chairs on the deck of the Titanic.
I suppose some courts will opt to give an iota of credit for the side that found AI hallucinations. The key there is that if that side has proceeded to tell the court, they have helped the court to be alerted to the matter. The court can then do a further inquiry and figure out that both sides have done likewise. Bringing up the topic does have some handiness, though there is still the unadorned fact that the side providing the alert had made use of AI hallucinations.
Let’s keep our eye on the ball.
Recent Case Highlights The Double Trouble
A recent court case has brought to the news and social media this zany aspect of both sides of a case making use of AI hallucinations.
In the U.S. District Court for the Northern District of Mississippi, Aberdeen Division, Civil Action, No. 1:24-CV-218-SA-RP, the case of Tom Withers III v City of Aberdeen, there is a Sanctions Order that was filed on June 8, 2026, and made these points (excerpts):
- “This case presents the Court with an unusual scenario — attorneys for both litigants engaged in similar sanctionable conduct.
- “As noted, Williams admitted to using AI to conduct legal research and incorporating the output into her respective filing without verification. By the same token, Wilson admitted to using generative AI to draft her respective filing without verifying the legal authority incorporated into the generated draft.”
- “Their practice of blindly relying on technology resulted in the hallucinatory citations contained in their respective filings.”
I believe this might be the first such court case to encounter this circumstance. Well, hold on, there is always the possibility that it has been happening, but nobody realizes that it has occurred. For example, suppose a case involves both sides using AI hallucinations, neither side detects it, and the court doesn’t detect it either. This would suggest that a court case and legal filings could easily be sitting out there that have indeed crossed the Rubicon; we just don’t know that it has happened.
Maybe we should clarify that this is the first court case in which the matter has been encountered and brought to prominent light. There might be others, maybe many others, quietly residing in the law books. I realize that seems far-fetched because you would hope that either side or the court would have caught the situation. Isn’t that supposed to be how our system works? Sorry to say, it is indubitably believable that such instances exist.
Deciding What To Do
I want you to assume that you are a judge who has uncovered that AI hallucinations exist on both sides of a case that you have before you.
What would you do about it?
First, you would certainly want to ask both sides to provide an explanation about their side of the story. Will they each admit to their faultiness? What kind of explanation do they have to offer? Is the explanation credible? Do you think that one side or both sides might be lying or stretching the truth about how they faltered? And so on.
The crux is that a judge would not want to fly off the handle before they have surfaced more details. It is giving the essential benefit of the doubt to the respective sides. The other utility in this is that it will be easier to determine the repercussions by having each side lay out in their own words the story they wish to tell about how the situation came into being.
Keep in mind that whatever sanctions or penalties the judge opts to impose might be brought up on appeal. A judge that did not do their dutiful homework is going to get a lot less latitude. Obtaining the stories from both sides would be suggestive that the judge was seeking to be fair and balanced, and likewise, the selection of repercussions would presumably be based on those respective stories.
Not All Get The Same Repercussions
One gut instinct might be that both sides should get the same sanctions or penalties. Since both sides made use of AI hallucinations, it stands to reason that they should suffer the same consequences. This seems straightforward.
It is important to realize that the devil is in the details. You cannot broadly assume that both should have the same consequences. For example, suppose that one side used only one AI-hallucinated citation, while the other used twenty. Does it seem balanced to say that they both committed the same type of error? You might insist that doing this twenty times is worse than having done it simply once.
Using a counter alone can be misleading. The one AI hallucination might have been crucial to the core of the case at hand. The twenty AI hallucinations might have been immaterial. In that instance, you cannot seemingly treat the sides as equal. The side that built their case on an AI hallucination would presumably be in hotter water.
Numerous factors need to be applied.
A conscientious judge would consider the nature of the instances, the weight of the instances, the filings they were in, whether this had happened in more than one filing, whether the attorney had ever previously done something similar, etc. The additional consideration is that if a side has multiple attorneys representing the case, the consequences probably ought to be aligned with each attorney and not be a blanket condemnation for the set of attorneys.
I want to clarify that this doesn’t mean that if there are four attorneys on one of the sides, and one admits they did it, while the other three feign innocence, this does not let those three off the hook. As officers of the court, having a duty to their professional obligations, and so on, they cannot try to pin all the blame on a fellow attorney. They must pay the piper too, though it might be a lesser consequence. One supposes that if there is a senior attorney that tries to act blameless, maybe they ought to get hit with staunch repercussions, assuming that the others were more junior, because they should have known better. Maybe this applies, maybe not.
The rule of thumb is that each attorney caught up in the morass deserves a consequence appropriate to their role and standing in the matter, and it is not particularly a one-size-fits-all condition.
Bringing Down The Hammer
A wide array of consequences have already been used in cases involving one side that made use of AI hallucinations. Courts have opted to sometimes just admonish the side that did this and move on without any other stirring repercussions. Sometimes the court requires the side that used AI hallucinations to cover the specific costs of the other side and the court costs in dealing with the matter.
Courts have admonished and reported attorneys to professional boards. Judges have handed out sanctions and required the payment of special penalty fees. In rare instances, a court might directly impact the client of the side that did this, reaching beyond the attorneys at fault. Some are queasy about clients being undermined by the acts of their attorney when it comes to AI hallucinations. That might be a bridge too far.
With that full range of consequences being on the table, what should happen when both sides are found to be using AI hallucinations?
As noted, the court should seek to ensure that the penalty fits the misconduct, in the sense that the consequences are to be tailored to each side, and to each respective attorney, rather than being overly broadly applied. That’s not to say that it couldn’t be that both sides end up with the same consequences. This surely could be justifiable and sensible.
Knock Your Socks Off
I think you are going to be surprised to know what happened concerning the AI hallucinations consequences in the civil case cited above. You might want to sit down before finding out. Let me know when you are ready for the news flash.
Okay, here we go.
The court decided that the attorneys on both sides were no longer allowed to represent their respective sides of the case. They got kicked off the case. The case is stayed for a period of time to allow the selection of new attorneys. The plaintiff would especially need to find new attorneys if the case is to proceed forward; otherwise, all else being equal, the case will ultimately be dismissed.
The court also imposed other sanctions and penalties, which I won’t list in detail, so make sure to read the case if you’d like to see the full blaze of consequences.
Here is a snippet of the ruling (excerpts):
- “The Court finds that the following sanctions are warranted and necessary to deter similar conduct by members of the larger bar and that no lesser sanction will serve that necessary deterrent purpose, vindicate the Court’s interest, or otherwise rectify this misconduct.”
- “First, pursuant to this Court’s Local Rule 83.1(c) and (d)(8), Wilson’s and Williams’ pro hac vice admissions in this case are hereby REVOKED.”
- “Secondly, pursuant to Rule 11, Ridgeway and McClinton are hereby DISQUALIFIED from further participation in this case…”
The court provided an overview of the explanations that each side provided, along with the logic of the court in deciding upon the appropriate set of consequences. I’m sure this will be handy if there are any appeals. Plus, this case provides a useful historical precedent that can be potentially leaned into for any future such circumstances.
Other courts can look at how this was handled. Attorneys in the future that are caught in such a situation can also potentially use this case to support whatever story they wish to tell, including legally arguing the adjudicated consequences they should face.
Double Whammy And Double Trouble
I believe we will see more of these double troubles and double whammy situations.
Whether some attorneys are truly unaware of the existence of AI hallucinations, or think that it can’t happen to them, or roll the dice on it, there are going to be instances of both sides of the case making use of AI hallucinations. The interesting turn will be whether one side figures it out, and if so, whether they admit to their own failings while pointing fingers at the other side.
There is some legal strategy to be considered. If one side does detect that the other side has AI hallucinations, but then realizes they have done so too, maybe they will decide to keep quiet. The idea of tattling on the other side is going to get them mired in the same muck. They would end up tattling on themselves too. Perhaps it would be wiser to hope that the other side uses the same kind of logic and remains quiet, or that the other side doesn’t ultimately figure out that AI hallucinations are in their midst.
Meanwhile, there is a ticking clock that the court itself might ferret things out. The other complication is that if one side or both sides did detect the AI hallucinations and opted to hold their tongues, does that constitute a violation of their duty to the court and the system of justice? Playing hide-and-seek could be worse than fessing up.
It is all a high-stakes gambit. A final thought for now. Remember the insightful words of William Shakespeare: “Double, double, toil and trouble; Fire burn, and cauldron bubble.”
Wise considerations for attorneys in the era of AI.
