Sanction teardown · D. Colorado, USA · 2025-10-30

Nonnie Berg v. United Airlines, Inc. (1)

What happened

In D. Colorado, USA, a filing relied on an unnamed/unconfirmed AI tool to help draft legal argument. The court identified the following problems with the citations in that filing:

• Fabricated (Case Law)
  Plaintiff cited Salt Institute v. Leavitt, 440 F. Supp. 2d 240, 243 (D.D.C. 2006); court could not identify this citation in Westlaw and treated it as an AI-generated, non-existent opinion.
• Fabricated (Case Law)
  Plaintiff cited In re Parke-Davis Sec. Litig., 177 F.R.D. 246, 250 N.6 (E.D. Mich. 1997); court could not verify the citation and treated it as fabricated AI output.
• Fabricated (Case Law)
  Plaintiff cited Koch v. Koch Indus., Inc., 203 F.R.D. 479, 487 (D. Kan. 2011); court could not locate this citation in its database and treated it as a non-existent case likely generated by AI.
• Fabricated (Case Law)
  Plaintiff cited Hodgson v. Mason & Hangar-Silas Mason Co., 545 F. Supp. 1189, 1192 (D. Kan. 1982); court could not identify this citation and considered it an AI-fabricated opinion.
• Fabricated (Case Law)
  Plaintiff cited Smith v. Life Investors Ins. Co. of Am., 245 F.R.D. 580, 584 (D. Kan. 2007); court was unable to verify the citation and treated it as a fabricated case citation from AI-generated content.
• Fabricated (Exhibits & Submissions)
  Plaintiff submitted a medical report (ECF No. 108-3) dated April 27, 2020 that the court found to contain indicia of AI generation (generic language, markdown artifacts, incomplete license info) and could not verify the physician.

Which AI tool

an unnamed/unconfirmed AI tool. 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

Warning

Additional detail

In an earlier Report and Reccomendations, the court found that significant portions of the plaintiff's filings copied from an AI program included citations to cases that could not be identified in Westlaw and an apparent AI-generated medical report; the court struck the filings and instructed compliance with Rule 11 and practice standards.

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/973/Berg_v._United_Airlines_USA_30_October_2025.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).