AI does not eliminate the consequences of bad legal work
Sanctions remain contingent if AI-assisted briefing contains factitious citations
A court order in Clark v. Social Security, Commissioner of links AI-assisted briefing to a possible sanction, but does not itself impose one.
Assembled by AI Law Firm News Desk.

AI-generated image
On September 24, the court in Clark v. Social Security, Commissioner of ordered Rich to show cause why she should not be sanctioned, including with a fine, for including factitious citations in a plaintiff’s brief, according to the CourtListener RECAP docket. The filing says Rich used generative artificial intelligence in preparing the brief. It also says she personally addressed the court’s inquiry about the filing. The immediate event is therefore a show-cause order tied to the quality of the finished legal work, not a sanction already imposed.
That distinction makes the order a direct example of the position that AI does not eliminate the consequences of bad legal work. The court’s concern is not simply that a tool was used. It is that the brief contained factitious citations, and the possible response reaches the person responsible for filing it. AI can change how a brief is produced without changing the consequences attached to an inaccurate brief. For an AI-native legal practice, that is the relevant operating condition: a system can promise speed or scale, but the work still has to survive review by the tribunal receiving it.
The order proves that one court has required an explanation for a brief containing factitious citations after AI was used in its preparation. It does not prove that AI assistance alone is sanctionable, that every AI-assisted brief contains false citations, or that a sanction will follow in this matter. The CourtListener: AI and privilege opinions entry for Douglas v. Deutsche Bank National Trust Co., Published Order describes AI misuse as carrying consequences for attorneys, while the Legaltech News account of a separate New Jersey lawsuit describes an AI meeting tool as capturing and distributing information in unintended ways. Those items reinforce the broader point that AI use can create legal exposure, but they do not convert the Clark order into a general rule. The stronger reading would require a final order stating whether Rich is sanctioned, what sanction is imposed, and what conduct supports it. Until then, the consequence is possible and formally presented, not completed.
The next fact to watch is the court’s resolution of the show-cause process in Clark v. Social Security, Commissioner of. A final order imposing, declining, or narrowing a sanction will show whether the possible fine becomes an actual consequence and whether the court treats the factitious citations, the AI use, or both as the decisive conduct. We look for that order rather than a broader promise that review has been automated away: finished legal work remains accountable when its citations fail.
News and analysis, not legal advice.
Sources
- Clark v. Social Security, Commissioner of CourtListener RECAP dockets: generative AI and sanctions or hallucinations
- Douglas v. Deutsche Bank National Trust Co., Published Order CourtListener: AI and privilege (opinions)
- 'Strapping Young Man': Ex-Employee Sues Over Conversation Captured by AI Legaltech News (Law.com)