The Fifth Circuit Says Generative AI May Be a New Technology but the Same Sanctions Rules Apply

The US Court of Appeals for the Fifth Circuit opens its opinion of February 18, 2026 on an issue it calls "central to the ongoing discussions of the relationship between law and technology: the use of artificial intelligence ("AI") in the drafting of legal documents." It had ordered a lawyer to show cause why she should not be sanctioned for a brief with quotations, citations and assertions "that were not supported by the underlying case law."

Having read her responses, the court found that she "used artificial intelligence to draft a substantial portion, if not all, of her reply brief and then failed to verify the accuracy of the content generated," and that she "was not forthcoming in her response to the show-cause order." It ordered her to pay a sanction of $2,500 within 30 days. Parts of the opinion follow, not the whole, and no lawyer, firm or party it names is named here.

A Rule Not Adopted, and a Growing Problem

After reports of AI hallucinations in court filings, the court appointed a subcommittee of three judges to study the issue. The subcommittee drafted a rule that would have required lawyers and litigants who represent themselves to certify either "that no generative AI program was used to prepare any submitted document," or, if one was used, "that a human checked the AI-generated text for accuracy."

The court published the proposal and asked for comment. Many of the comments said such a rule was unnecessary, because existing rules already oblige counsel to give courts accurate information. The court followed its subcommittee and declined to adopt it, issuing instead a notice reminding counsel "of their obligation to review all filings for accuracy." It concluded "that existing rules were sufficient to deter misconduct related to generative AI use, without the need for a rule specific to generative AI."

AI-hallucinated case citations, the court says, "have increasingly become an even greater problem in our courts, and the problem shows no sign of abating." It says the problem "now often manifests as false quotes or statements of law attributed to real cases, rather than the more easily recognizable fake cases." "If it were ever an excuse to plead ignorance of the risks of using generative AI to draft a brief without verifying its output, it is certainly no longer so."

The rule the court chose not to adopt would have asked lawyers and litigants who represent themselves to certify, if they used a generative AI program, that a human checked its text. That check is the step the court found this lawyer had skipped.

The Answers to the Court

The court's order to show cause listed 16 instances of fabricated quotations and 5 more serious misrepresentations of law or fact. The lawyer first said she had relied "on publicly available versions of the cases," which she believed were accurate. The court found that answer "incredible on its face" and asked further questions.

Asked which sources had led to the errors, the lawyer named several well-known legal databases. The court says one of them "does not contain summaries of legal cases that could contain inaccurate quotes." For three others it reviewed their summaries of each case cited in the brief, and found none of the problematic quotations and propositions in them. The remaining two, it says, are generative AI products. She had not mentioned AI in her first response. When asked directly, she admitted to using it, and the court did not find it credible that she used it only to help organize her arguments and break up long paragraphs. "Even when asked directly," the court says, the lawyer "did not explain the steps she took to verify the factual assertions in her brief, so the court concludes that she took none."

Had the lawyer accepted responsibility and been more forthcoming, the court says, "it is likely that the court would have imposed lesser sanctions."

The Same Rules, and What Ethical Use Requires

"Modern generative AI may be a new technology, but the same sanctions rules apply, and the rules we have are well equipped to handle these types of cases," the opinion says. It rests the sanction on two of them: the rule that lets a court of appeals discipline an attorney for "conduct unbecoming a member of the bar or for failure to comply with any court rule," and the court's "inherent power to impose sanctions for abuse of the judicial process." Submitting a brief "riddled with fabricated quotations and assertions," it says, is such an abuse.

The court does not dispute that generative AI "can be helpful if done properly and carefully." To use it ethically, it says, quoting another court, a lawyer must "ensure that the legal propositions and authority generated are trustworthy."