Sanctions For Misuse Of AI In Court Submissions Increasing Sharply: Another Cautionary Tale

AI Court Filing Sanctions: A $60K Wake-Up Call for Firms

While we have previously reported an increase in fines associated with the misuse of generative artificial intelligence in court submissions, new sanctions handed down in recent months a significantly more serious than in prior situations, indicating courts across the country are increasingly losing their patience. 

Illinois Court Sanctions Firm Over AI Case Citations 

In December 2025, an Illinois circuit court sanctioned a law firm $50,000 and the individual attorney another $10,000, for the inclusion of false case citations and quotations (known as AI hallucinations) in post-trial submissions to the court. Plaintiff’s counsel responded to the defendant’s briefing, noting that as best as they could determine, one of the key cases cited in the defendant’s post-trial briefing “does not exist... [but] appears to have been invented out of thin-air.” On reply, defense counsel apologized “for this serious lapse in professionalism,” and maintained that many attorneys had assisted lead counsel in preparing the briefing and an “exhaustive investigation” pointed to one attorney who used Al technology and “failed to verify the Al citation before including the case and the surrounding sentence describing its fictitious holding.” Defense counsel’s firm advised the Court that it had undertaken to re-educate its attorneys on proper use of AI technology and advised it was available for more explanation as needed.

How the AI Hallucination Escalated to a Sanctions Hearing

The Court then set a hearing date and ordered all attorneys involved in the submission to be present to explain the misconduct. The drafting attorney, a partner with eleven years of experience, advised the Court that she had used Chat GPT to generate the text. She claimed that she knew AI technologies can create hallucinations but did not understand that it could generate fake caselaw. She further reported that she knew her work would be—and in fact was—reviewed by other partners working the case, including lead counsel, before the briefing was submitted to the court. Lead counsel argued it was just one citation of 58 cases cited in the brief and stated that he did not consider case review something that was a part of his final review before filing. Notably, while an earlier policy of the firm required IT department and firm leadership approval to use artificial intelligence in briefing, the firm policy in place at the time of the issue required only that any attorney using artificial intelligence in briefing “was required to meticulously review any output received through Al for content validity and other concerns.”

Even more, various attorneys at the firm (of roughly 440), including the drafting attorney, had previously published articles about the dangers of using AI without verification. 

The same drafting attorney was previously sanctioned by another court for filing a motion to dismiss that included at least 12 invalid legal citations arising out of her use of AI, particularly Chat GPT, to research and write court submissions.

Fourteen Fabricated Holdings: The Scope of the AI Misuse

At the conclusion of the hearing, the Court granted Plaintiffs permission to file a motion for sanctions, which they did. Contrary to lead defense counsel’s argument that there was only the one false case citation, Plaintiffs’ counsel identified fourteen instances where cases cited by defense counsel did not stand for the propositions cited. As the Court noted, these were not nuanced differences, but “examples of non-existent quotes and fabricated holdings.” Plaintiffs’ motion also identified AI hallucinations in two prior filings by defendants in the same case.

What the Court's Sanctions Order Actually Penalized

In its order of sanctions, the Court characterized the conduct as a “serious failure” of the attorneys’ obligation to present truthful, accurate information to the court—particularly legal authority. The Court further emphasized its focus was “not on the misuse of artificial intelligence to conduct unreliable legal research and drafting,” but “the inexcusable submission of false authority and factual arguments to the Court, the subsequent misrepresentations about the extent of the improper conduct, and the failure to take prompt responsibility for errors once discovered.” Importantly, the Court noted that the “most serious sanctionable conduct consists of actions taken after the attorneys had time to consider the consequences of submitting false statements of law and facts to the court, and had time to discover and disclose the full extent of the errors in citations and in factual assertions.” 

Finally, the Court reminded counsel that the “obligations on officials of the court at issue … precede by centuries the age of electronic research and artificial intelligence.” See Jordan v. Chi. Hous. Auth., No. 2022 L 000095 (Ill. Cir. Ct. Dec. 5, 2025).

Key Takeaways for Law Firm AI Policy and Candor Obligations

There are many takeaways here. The first being that most, if not all, states maintain policies specific to artificial intelligence (such as the Illinois Supreme Court Rule 137 at applied here) that may form the basis for sanctions in the event of any misuse of AI technology. The second takeaway is the need for full and complete disclosure in the event of such an error. Much of what this Court focused on in its order awarding sanctions was the failure of counsel to fully identify and disclose all errors and to take direct responsibility for those errors (particularly the lead attorney who signed the pleadings and was responsible for the content of those pleadings). Third, this case highlights the expectation of this Court (and likely many others) that attorneys must adhere to their obligations of candor to the Court. And finally, this order provides further support for the need of all law firms to have (and enforce) a comprehensive internal policy regarding the use of AI in legal research and writing.