Landberg v. City of New York — an AI-fabricated appellate brief drew $10,500 in sanctions
A New York appellate court (Second Department) sanctioned attorney Michael Sanders $8,000 and his law firm $2,500 — $10,500 total, to the Lawyers' Fund for Client Protection — after an appellant's brief prepared with generative AI cited three nonexistent cases, fabricated Court of Appeals quotations contrary to the actual law, and misrepresented real cases (2026 NY Slip Op 03935, decided June 23, 2026).
| Metric | Before | After |
|---|---|---|
| $8,000 sanction on attorney Michael Sanders, payable to the Lawyers' Fund for Client Protection of the State of New York | ||
| $2,500 sanction on the Law Offices of Michael S. Lamonsoff, PLLC, to the same fund ($10,500 combined) | ||
| Three nonexistent cases cited, plus fabricated Court of Appeals quotations contrary to actual law and misrepresentations of real cases | ||
| Sanction imposed under 22 NYCRR 130-1.1 (frivolous conduct); possible Attorney Grievance Committee review noted | ||
The problem
An attorney’s court filing is only as trustworthy as its citations. Generative-AI research tools will fabricate case names, invent quotations, and misstate holdings with complete fluency — and an appellate brief that repeats them without verification wastes the opposing party’s time, burdens the court, and can deprive the client of arguments grounded in real precedent.
Verification status: PENDING — checker-graduated at confidence 1.000, awaiting HUMAN client-confirmation. Not verified; no green badge until a human confirms the public record is current and unamended.
What was built
In Landberg v City of New York (Appellate Division, Second Department, Docket 2025-02380, Index No. 509670/18), the plaintiff’s appellant’s brief — submitted by attorney Michael Sanders of the Law Offices of Michael S. Lamonsoff, PLLC — was, in the court’s own words, “prepared with the assistance of generative artificial intelligence (hereinafter GenAI), containing citations to nonexistent cases.” The underlying case was a sidewalk trip-and-fall over a loose brick in a tree well; the AI-drafted brief was the vehicle for the sanctioned conduct, not the merits.
The brief “cited to several nonexistent cases, misstated the law, and misrepresented the holdings of real cases.” It relied on three fictitious cases — Xiang Fu Ji v City of New York, Hausser v Giunta (88 AD3d 969), and Lack v Lack — and quoted a fabricated New York Court of Appeals passage that said the opposite of the actual law. Sanders acknowledged submitting “a brief … unknowingly with three fabricated citations,” which he believed “originated during the AI-assisted portion of [his] supplemental research which [he] negligently failed to verify.”
The outcome
In a Decision and Order on Motion decided June 23, 2026 (2026 NY Slip Op 03935), the court, acting under 22 NYCRR 130-1.1, held: “we deem it appropriate to impose a monetary sanction on Sanders in the sum of $8,000” and “we deem it appropriate to impose a monetary sanction on the law firm in the sum of $2,500.” Sanders was “directed to pay a sanction in the sum of $8,000 to the Lawyers’ Fund for Client Protection of the State of New York,” and the firm $2,500 to the same fund — $10,500 combined. The court added that “[w]hether disciplinary action is also warranted may be a matter for the Attorney Grievance Committee.”
The court also faulted Sanders’s candor: he initially told the panel the authorities came from Lexis, Westlaw, a book or a prior case, and only later admitted in an affirmation that “he should have said that the cases were hallucinated by artificial intelligence.”
Conflation guards. The opinion names no AI product (do not attribute it to any brand); the $250 figure in the opinion is a different case (Matter of Julien v Arthur) cited by comparison; and the sidewalk-injury merits are unrelated to AI.
Path to green
This is a court-origin record, not a vendor claim: every figure is quoted verbatim from the published appellate opinion, which is byte-tied to an archived copy. No client-confirmation call is required. The only human step to a green badge is an editorial confirmation that the published opinion (2026 NY Slip Op 03935 on the official New York State Law Reporting Bureau site) is authentic, current, and undisturbed by any reargument — a decision reserved to a human, never the war-room.
Generative AI (unnamed) used for legal research/drafting22 NYCRR 130-1.1 (New York frivolous-conduct sanctions)
- Status
- pending
- Method
- Court-origin: published appellate opinion (2026 NY Slip Op 03935) quoted verbatim and byte-tied to Wayback via a byte-identical S3 re-host; corroborated by two independent newsrooms (BridgeTower/NY Daily Record and Mealey's/LexisNexis on the figures; 404 Media on the fabrications).
- Provider
- Unnamed generative-AI research tool ('one of the free ones available to the public')
- Client
- Supreme Court of the State of New York, Appellate Division, Second Judicial Department · Courts / legal (AI-governance honest-negative)
- Disclosure
- named