What was hallucinated

Fabricated: Case Law | CHA relied on a non-existent case as controlling authority for admitting alternative-source evidence; plaintiffs identified it as fabricated and CHA conceded it was produced by an attorney using AI. || Fabricated: Case Law | Another non-existent case cited in CHA filings identified by plaintiffs and attributed by the firm to AI-assisted drafting. || Misrepresented: Case Law | CHA misstated the holding of Barry v. Owens-Corning Fiberglass Corp., portraying it as affirming remittitur when the Illinois Supreme Court did not vacate the award. || Misrepresented: Case Law | CHA represented Hollis v. R. Latoria Construction as supporting remittitur where the Supreme Court actually affirmed and remanded for a new trial based on inadequacy of damages. || Misrepresented: Case Law | CHA mischaracterized Johnson v. Mers as involving a trial court reduction of pain-and-suffering awards; the case actually affirmed summary judgment for the employer and did not involve that trial scenario. || Misrepresented: Case Law | CHA quoted Sears v. Rutishauser for a principle about speculative damages requiring reduction, but the court held a different point regarding cross-examination of a treating physician; the quotation/point was misapplied. || False Quotes: Case Law | CHA included a quoted passage attributed to Thompson v. Gordon that the opinion does not contain; the court found the quote absent and the proposition unsupported by that case.

Details

(Motion for sanctions available here.)"The court’s focus here is not the misuse of artificial intelligence to conduct unreliable legal research and drafting. It is 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. The obligations on officers of the court at issue here precede by centuries the age of electronic research and artificial intelligence.The failures to meet those obligations do serious damage to the respect for the legal profession, and they merit sanctions. 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.[...]Artificial Intelligence is not the cause of bad legal practice. Lawyers performed their obligations well and performed their obligations poorly before Al, before electronic research platforms, before on-line publication of case law, and before the development of the West Key Number System or Shepard’s indexes.Submission of false legal citations and demonstrably false factual claims pose a grave threat to the judicial branch. People are skeptical of institutions, and the legal profession is not exempt. We are duty-bound to attend to the integrity the courts so that close scrutiny reveals a model of honesty, accountability, and truth-seeking.The authority of the courts relies on public confidence that rulings are just and are grounded in the law, not on the whims of judges. “[A] lawyer should further the public’s understanding of and confidence in the rule of law and the justice system because legal institutions in a constitutional democracy depend on popular participation and support to maintain their authority.” (IRPC Preamble, par. 6) Officers of the court cannot become comfortable with careless or deliberate misrepresentation of facts or the law."

Sanction teardown · CC Illinois, USA · 2025-12-05

Shana Jordan, et al. v. Chicago Housing Authority et al.

What happened

In CC Illinois, USA, a filing relied on ChatGPT to help draft legal argument. The court identified the following problems with the citations in that filing:

Fabricated (Case Law)

CHA relied on a non-existent case as controlling authority for admitting alternative-source evidence; plaintiffs identified it as fabricated and CHA conceded it was produced by an attorney using AI.
Fabricated (Case Law)

Another non-existent case cited in CHA filings identified by plaintiffs and attributed by the firm to AI-assisted drafting.
Misrepresented (Case Law)

CHA misstated the holding of Barry v. Owens-Corning Fiberglass Corp., portraying it as affirming remittitur when the Illinois Supreme Court did not vacate the award.
Misrepresented (Case Law)

CHA represented Hollis v. R. Latoria Construction as supporting remittitur where the Supreme Court actually affirmed and remanded for a new trial based on inadequacy of damages.
Misrepresented (Case Law)

CHA mischaracterized Johnson v. Mers as involving a trial court reduction of pain-and-suffering awards; the case actually affirmed summary judgment for the employer and did not involve that trial scenario.
Misrepresented (Case Law)

CHA quoted Sears v. Rutishauser for a principle about speculative damages requiring reduction, but the court held a different point regarding cross-examination of a treating physician; the quotation/point was misapplied.
False Quotes (Case Law)

CHA included a quoted passage attributed to Thompson v. Gordon that the opinion does not contain; the court found the quote absent and the proposition unsupported by that case.

Which AI tool

ChatGPT. Note: Charlotin's public database records tool attribution only where a court order, brief, or reporting on the matter states it explicitly; "unidentified" or "implied" means the record indicates AI use but does not name a specific product — we do not guess.

Outcome

Motion partly struck; Monetary sanctions (monetary penalty: 59500 USD.)

Additional detail

(Motion for sanctions available here.)"The court’s focus here is not the misuse of artificial intelligence to conduct unreliable legal research and drafting. It is 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. The obligations on officers of the court at issue here precede by centuries the age of electronic research and artificial intelligence.The failures to meet those obligations do serious damage to the respect for the legal profession, and they merit sanctions. 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.[...]Artificial Intelligence is not the cause of bad legal practice. Lawyers performed their obligations well and performed their obligations poorly before Al, before electronic research platforms, before on-line publication of case law, and before the development of the West Key Number System or Shepard’s indexes.Submission of false legal citations and demonstrably false factual claims pose a grave threat to the judicial branch. People are skeptical of institutions, and the legal profession is not exempt. We are duty-bound to attend to the integrity the courts so that close scrutiny reveals a model of honesty, accountability, and truth-seeking.The authority of the courts relies on public confidence that rulings are just and are grounded in the law, not on the whims of judges. “[A] lawyer should further the public’s understanding of and confidence in the rule of law and the justice system because legal institutions in a constitutional democracy depend on popular participation and support to maintain their authority.” (IRPC Preamble, par. 6) Officers of the court cannot become comfortable with careless or deliberate misrepresentation of facts or the law."

How Citation Safe would have caught this

Citation Safe runs three deterministic layers before a brief is filed: (1) does the citation exist against CourtListener's database of published opinions, (2) if quoted, does that exact language appear in the source, (3) does the cited case actually support the proposition it is cited for. Fabricated case citations fail Layer 1. Fabricated or misattributed quotations fail Layer 2 even when the underlying case is real. Misrepresented holdings — a real case cited for a proposition it does not support — are the target of Layer 3. None of these checks involve asking another language model whether the citation looks right; they are lookups and text-matches against the actual source, which is why a hallucinated citation has to survive a direct lookup against the authoritative source — not another model's opinion — to earn a VERIFIED stamp; our measured false-verify rate is published live at /quality.

Check a brief before you file it → · See our live false-verify rate

Source: https://www.damiencharlotin.com/documents/1159/Jordan_v._CHA_USA_5_December_2025.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).