Noland v. Land of the Free: California's first published opinion on AI-fabricated citations ($10,000 sanction)

The problem

Sylvia Noland appealed a summary judgment against her wage-and-hour claims (source). What set the appeal apart, the Court of Appeal wrote, was that “nearly all of the legal quotations in plaintiff’s opening brief, and many of the quotations in plaintiff’s reply brief, are fabricated” (source). The quotes counsel attributed to published cases did not appear in those cases; some cited cases did not discuss the cited topics, and a few did not exist (source). “These fabricated legal authorities were created by generative artificial intelligence (AI) tools that plaintiff’s counsel used to draft his appellate briefs,” the court found (source). Independent reporting put a number on it: 21 of 23 case quotations in the opening brief were fabricated, along with many more in the reply brief (source).

What was built

This is an AI-governance honest-negative, not a deployment win: the “system” is generative AI used as an unverified legal-drafting tool. At oral argument, counsel described the workflow, saying he “wrote initial drafts of the briefs, ‘enhanced’ the briefs with ChatGPT, and then ran the ‘enhanced’ briefs through other AI platforms to check for errors” (source). Counsel “admitted that he did not read the ‘enhanced’ briefs before he filed them” (source). The court published the opinion deliberately: “no California court has addressed this issue. We therefore publish this opinion as a warning” (source).

The outcome

The figure that gets quoted: a $10,000 sanction on appellant’s counsel Amir Mostafavi. Affirming the judgment, the court imposed the sanction for prosecuting a frivolous appeal (source). Because counsel represented that his conduct was unintentional and expressed remorse, the court set a “conservative sanction of $10,000,” payable “to the clerk of this court within 30 days of the filing of the remittitur” (source). The court also directed Mostafavi to serve a copy of the opinion on his client and directed the clerk to forward a copy to the State Bar (source).

What the court declined to do is itself the news. It refused to award the sanction to opposing counsel, explaining that “while we have no doubt that such sanctions would be appropriate in some cases, in the present case respondents did not alert the court to the fabricated citations” (source). Legal Newsline reported the sanction firsthand, describing a Los Angeles plaintiff’s attorney fined $10,000 for an opening brief “riddled with AI-generated fabrications” (source).

How this was verified

The primary source is the signed, published opinion of the California Court of Appeal, Second District, Division Three (B331918), which is self-authenticating: the $10,000 sanction, the 30-day-after-remittitur deadline, the client-service directive and the State Bar referral are stated verbatim in it, and every quotation above was re-checked against the published text on 2026-08-17 (source). The 21-of-23 fabrication count is not stated in that phrasing in the opinion itself; it comes from independent legal reporting (LawSites and Legal Newsline reading the opinion), and is presented with that attribution (source). The honest limit: the downstream State Bar disciplinary outcome is not part of this record and was not independently re-confirmed this session, so the badge rests only on the court-adjudicated facts above, not on any final discipline.

Related case files

The nearest sibling is Torres Campos v. Munoz, a later published California appellate opinion that sanctioned counsel for the same conduct, showing the warning Noland issued did not stop the pattern in the same courts. Kruse v. Karlen is the closest match on the facts: an appellate brief where nearly every citation was fabricated, drawing the same $10,000 frivolous-appeal sanction. Barber v. Morawa is the Michigan counterpart, another state’s first published opinion on AI-fabricated citations issued in the same deliberate “warning” posture.

Sources

Tiers: Tier 1 primary (the court’s own signed opinion); Tier 2 independent reputable press naming the parties. Figures re-checked live on 2026-08-17.

1. California Court of Appeal, 2d Dist., Div. 3, “Noland v. Land of the Free, L.P.,” B331918, filed 2025-09-12 (Tier 1, the signed published opinion; reproduced by FindLaw and re-checked verbatim). https://caselaw.findlaw.com/court/crt-app-sec-dis-cal-div-thr/117694279.html
2. California Courts, official slip opinion PDF (Tier 1 primary host; byte-tied to a 2025-09-13 Wayback capture). https://www4.courts.ca.gov/opinions/documents/B331918.PDF
3. Legal Newsline (Southern California Record), “CA court hits lawyer with $10K fine for AI citations, issues warning,” 2025 (Tier 2, independent legal newsroom reporting the sanction firsthand). https://www.legalnewsline.com/south-california-record/ca-court-hits-lawyer-with-10k-fine-for-ai-citations-issues-warning/article_4c991205-90b7-4bee-86d2-f25fbee8991e.html
4. LawSites (Bob Ambrogi), “A new wrinkle in AI hallucination cases,” 2025-09 (Tier 2, independent; source for the 21-of-23 count and the fee declination, re-confirmed 2026-08-17). https://www.lawnext.com/2025/09/a-new-wrinkle-in-ai-hallucination-cases-lawyers-dinged-for-failing-to-detect-opponents-fake-citations.html