A brief can draw sanctions if AI-generated citations go unchecked

CourtListener RECAP dockets reports that, in Clark v. Social Security, Commissioner of, a court ordered Rich to show cause after a brief prepared with generative artificial intelligence included factitious citations. The docket was filed on September 24, 2026, and says the order included the possibility of a fine. The order therefore moves the dispute beyond whether AI was used: it puts the accuracy of the finished legal work before the court and attaches a possible sanction to the filing.

That is the point for AI-native legal work. AI can change how legal work is produced without changing who answers for what reaches a court. A system that produces a brief with false citations has not removed the consequence of bad legal work; it has made verification the condition for delivering that work safely. The Clark order does not show that AI-assisted work is inherently defective. It does show that a court can treat an AI-assisted filing as a legal filing, with the same exposure to scrutiny that applies to any other brief. The relevant measure is not whether a tool saves time or produces fluent prose. It is whether the citations and propositions in the finished work survive checking before submission.

The event proves a narrower point than a final sanctions ruling would. CourtListener RECAP dockets says the court ordered a show-cause response; it does not say that a fine was imposed, what amount a fine would take, or whether the court found that every challenged citation came from AI. It also does not establish that all AI-assisted briefs create the same problem. A separate CourtListener: AI and privilege (opinions) entry for Douglas v. Deutsche Bank National Trust Co., Published Order describes misuse of AI and its consequences for attorneys, but that description does not turn the Clark matter into a general rule. The stronger reading would require a later order explaining the citations, the responsibility for them, and the court’s chosen consequence. Until then, Clark is one filing that triggered judicial scrutiny, not proof that AI-native delivery has failed or that verification is optional.

The next fact to watch is the court’s ruling on the Clark show-cause order. A sanctions order that identifies the factitious citations, ties them to the filing process, and imposes a fine would confirm that AI use does not displace responsibility for the final brief. A response that resolves the matter without a sanction would narrow the lesson, but would not erase the need to determine whether the citations were checked. We should score the claim on that concrete ruling, not on the presence of AI in the workflow alone.