Skip to content

Incident Tracker / United States (federal)

Benjamin v. Costco Wholesale Corp., No. 2:24-cv-7399 (LGD) (E.D.N.Y.)

·US District Court, Eastern District of New York·US-FEDFine

In the US District Court for the Eastern District of New York, Magistrate Judge Lee G. Dunst fined attorney Victoria Hovsepyan $1,000 under Rule 11 after she used the AI platform ChatOn to draft a reply on a remand motion that cited five fake cases; the court found subjective bad faith but declined any disciplinary referral.

Court
US District Court, Eastern District of New York
Date
2025-04-24
Who used the AI
Lawyer
AI tool
ChatOn
Conduct
Plaintiff's counsel Victoria Hovsepyan, pressed for time after a paralegal's draft proved unusable, used the AI app ChatOn to draft a reply in support of a motion to remand and filed it as a sworn affirmation without reading the cases; five cited cases were fake. Magistrate Judge Lee G. Dunst found a Rule 11 violation and subjective bad faith.
Outcome
Fine
Penalty
US$1,000
Regulatory outcome
No separate regulatory disposition recorded
Status
Verified against listed sources
Last checked
2026-09-04

What was fabricated or misused

  • Klein v. E. I. Du Pont de Nemours & Co., 406 F.2d 1004, 1006 (2d Cir. 1969) — does not exist; the citation leads to Spartans Indus., Inc. v. N.L.R.B.
  • Gordon v. N.Y. Cent. R.R. Co., 202 F. Supp. 2d 290, 292 (W.D.N.Y. 2002) — does not exist; the citation leads to Fournier v. Erickson
  • Mitchell v. JCG Industries, 2010 WL 4975563, at *3 (N.D. Ill. Dec. 1., 2010) — does not exist; the citation leads to People v. Castillo
  • Hollander v. Sweeney, 2005 WL 19904045, at *3 (S.D. Ind. Aug. 17, 2005) — does not exist; the Westlaw citation is a Federal Taxes Weekly Alert page
  • Davis v. S. Farm Bureau Cas. Inc. Co., 2019 WL 3452601, at *2 (W.D. Ark. July 31, 2019) — does not exist; the Westlaw citation describes an Ohio sales-tax certificate revision

What the court said

“None of that happened here. In this case, Plaintiff’s counsel, Victoria Hovsepyan, cited fake cases to the Court in her reply in support of a motion to remand.”
US District Court, Eastern District of New York, p. 2
“Plaintiff’s counsel concedes that these cases do not exist and she instead generated the citations through the use of an AI platform known as ChatOn.”
US District Court, Eastern District of New York, p. 6
“The Court is compelled to conclude this conduct rises to the level of subjective bad faith.”
US District Court, Eastern District of New York, p. 16
“A penalty of $1,000 is imposed on Ms. Hovsepyan.”
US District Court, Eastern District of New York, p. 17

Timeline

  1. Benjamin commenced the negligence action in New York state court.
  2. Judge Choudhury set the remand-motion briefing schedule.
  3. Counsel filed the reply affirmation containing five fake cases generated by ChatOn.
  4. Court ordered counsel to show cause or provide copies of the cited cases.
  5. Court held the sanctions hearing on the order to show cause.
  6. Court imposed a $1,000 Rule 11 penalty and client-notice requirement.

Why this case matters

Benjamin sits with Bevins v. Colgate-Palmolive and Dehghani v. Castro in the April 2025 federal wave of lawyer-filed hallucinated citations, but it adds a different finding. Judge Dunst found subjective bad faith because counsel signed a sworn affirmation after doing nothing to confirm ChatOn's cases, even though she quickly admitted the problem and took AI CLE courses. The sanction was lower than Mata v Avianca because the court credited her candour and did not find later deception.

Practice note

Before a reply is filed, counsel should open every cited decision and read the cited proposition in a legal database. A deadline problem should be handled by seeking more time, not by substituting an unreviewed AI draft. Swearing to a filing requires personal knowledge that its legal citations exist and support the argument.

Primary sources for this incident

Related regulation records for this incident

No related regulation record has been linked to this incident yet.

Similar incidents on the record

  • Chirece C. v. Commissioner of Social Security, No. 25-11554 (E.D. Mich. 2026)2026-09-24 · United States District Court for the Eastern District of Michigan · FineIn Michigan federal court, Magistrate Judge Elizabeth Stafford fined a Social Security claimant's lawyer $500 for AI-generated fictitious citations she admitted not checking. The same lawyer had been sanctioned in July…
  • Booker v. The Kroger Co., No. 1:26-cv-02006-SDG (N.D. Ga.)2026-08-28 · US District Court for the Northern District of Georgia, Atlanta Division · FineThe US District Court for the Northern District of Georgia fined employment plaintiff's lawyer Christopher Johnson USD 8,000, payable to the court's sanctions fund, for citing four nonexistent or misrepresented authorit…
  • Adams v. Matrix Providers, No. 1:23-cv-01996 (D. Colo.)2026-08-27 · US District Court for the District of Colorado · FineThe US District Court for the District of Colorado fined counsel Ms Pearson USD 1,000 after her brief cited a nonexistent case and attributed a quotation to Justice Scalia from an opinion dated 2020, four years after hi…
  • Hill v. Foundation Media LLC, No. 1:25-cv-05947 (AS) (S.D.N.Y. 2026)2026-09-29 · United States District Court for the Southern District of New York · WarningIn New York federal court, Judge Arun Subramanian ordered defence counsel to explain AI-affected quotations in a response brief, took no further action, and called the episode a wakeup call. Counsel's 1 October letter d…

Cite this record

SafeLegalAI Global Legal AI Incident Tracker, “Benjamin v. Costco Wholesale Corp., No. 2:24-cv-7399 (LGD) (E.D.N.Y.)”, safelegalai.com/tracker/benjamin-v-costco (accessed 2026-10-07). Data: CC BY 4.0.

Report an issue with this record →