Artificial Intelligence & Robotics

More courts are coming down on ‘non-offending counsel’ for AI missteps

GettyImages-AI words and gavel

Courts have since issued decisions with increasingly “stronger language” when admonishing “non-offending counsel,” says April Dawson, a 2026 ABA Journal Legal Rebel. (Photo by Alexander Sikov/Getty Images)

Amid the proliferation of cases involving artificial intelligence-generated hallucinations, more judges are expressing frustration not only at the attorneys who make the mistakes but at opposing counsel for not pointing it out.

In the past year, courts have admonished attorneys for failing to identify and report fake citations in their opponents’ court filings. In at least two cases, judges refused to award attorney fees or grant relief to counsel who didn’t bring AI-induced errors to their attention.

April Dawson, the associate dean of technology and innovation at the North Carolina Central University School of Law, began tracking this trend after the California Court of Appeal, Second Appellate District issued its decision in Noland v. Land of the Free. Last September, the court imposed a $10,000 sanction on the plaintiff’s counsel for filing briefs with fabricated citations and quotations, but it also declined to award attorney fees to the respondent’s counsel despite finding the appeal frivolous.

“While we have no doubt that such sanctions would be appropriate in some cases, in the present case, respondents did not alert the court to the fabricated citations and appear to have become aware of the issue only when the court issued its order to show cause,” the court said in its Sept. 12 opinion.

April DawsonApril Dawson. (Photo by Rick Crank)

Courts have since issued decisions with increasingly “stronger language” when admonishing “non-offending counsel,” says Dawson, a 2026 ABA Journal Legal Rebel.

In November, the Hennepin County District Court in Minneapolis said in Nuvola v. Wright that it was “troubling” that a lawyer did not check his opponent’s case citations, which turned out to be nonexistent. The court said it “should not be left as the last line of defense against citations to fictional cases in briefs,” but decided against imposing sanctions on the lawyer “as he did not cite any nonexistent cases.”

However, counsel are obligated “to respond to each other’s arguments, including completing a basic cite-check of the cases cited by the other side,” the court also noted.

“Courts are becoming increasingly more frustrated,” Dawson says. “When you have lawyers who are governed by the rules of professional responsibility, one, not understanding the technology, and two, not verifying cases they are citing, courts are going to continue to admonish them and levy penalties at greater and greater amounts.

“But I think they will also continue to call out opposing counsel, who themselves did not check cases, because we’re all aware now,” she adds.

‘A matter of perceived competence’

Lawyers filing documents with the court are required to ensure they are supported by evidence and warranted by existing law or by nonfrivolous arguments under Rule 11 of the Federal Rules of Civil Procedure. Meanwhile, Rule 3.3 of the ABA Model Rules of Professional Conduct states that a lawyer should not make false statements or offer evidence they know to be false to the court.

Thomas Wilkinson.

While these requirements may impact lawyers who file briefs with AI hallucinations, Thomas Wilkinson says they shouldn’t apply to lawyers who haven’t signed the filings or made inaccurate representations to the court.

“It’s a very interesting trend, and certainly a concern, but somewhat of a slippery slope to impose sanctions on lawyers whose own submissions are accurate,” says Wilkinson, a member of Cozen O’Connor in Philadelphia and of the ABA Standing Committee on Ethics and Professional Responsibility.

While lawyers may not be sanctioned for missing AI-related errors in opponents’ briefs, they also may not receive requested relief if the court is “disappointed in their lack of due diligence,” Wilkinson adds.

The U.S. District Court for the Northern District of Mississippi and the 7th U.S. Circuit Court of Appeals in Chicago are among other courts that recently faulted lawyers for not catching fake citations in their opposing counsel’s filings and reiterated their ethical duties to flag AI misuse when they see it.

The California Court of Appeal, Fourth Appellate District in March handed down even stricter penalties after counsel and the trial court itself did not detect AI hallucinations. The case arose out of family law matter in which a man, Joan Pablo Torres Campos, sued his former domestic partner, Leslie Ann Munoz, for shared custody and visitation of their pet dog. Munoz’s lawyer had cited two fake cases in her letter declining the request, to which Torres’ lawyer responded, but did not point out the hallucinations. Torres’ lawyer then submitted a proposed order and findings of fact that cited the cases as if they were real, and the trial court adopted them.

Torres appealed and claimed the trial court had erred by citing and relying on the fake cases, to which the appeals court agreed. But the appeals court also held that Torres had forfeited that claim because his counsel had not only failed to object to or notify the trial court of the fabricated citations but had drafted and submitted a proposed order containing them.

“Although we have discretion to overlook the forfeiture, we decline to do so because it would effectively excuse [the appellant]’s own counsel’s breach of her duty to the court in submitting the proposed order without verifying the citations,” the appeals court said.

In the past, lawyers may have only double-checked citations in their opponents’ briefs if they “seemed off or questionable, or not cited for routine propositions,” Wilkinson says. But now, that’s no longer acceptable as judges are raising the bar for due diligence.

“It’s a matter of perceived competence,” Wilkinson says.

Dawson agrees, contending that lawyers are obligated to zealously advocate for their clients, which includes carefully reading cases their opponents rely on in arguments.

“This is requiring counsel to continue to respond to all arguments in all cases, and by doing that, you should be able to identify cases that just don’t exist at all,” she says.

See also:

Lawyer can’t represent plaintiffs in morgue scandal because of past AI errors, judge says