A Kansas federal judge fined five attorneys $12,000 for AI-fabricated citations — and named the failure, not the AI

Verification status: pending — checker-graduated (confidence 1.000), awaiting HUMAN client-confirmation. NOT verified, NOT green.

The problem

In a patent-infringement suit over technology for modifying the display of a cursor image on a website (Lexos Media IP, LLC v. Overstock.com, Inc., D. Kan. No. 22-2324-JAR), plaintiff Lexos’s July 7, 2025 filings opposing Overstock’s summary-judgment and Daubert motions were riddled with fabricated legal authority. The opposition to Overstock’s motion to exclude Lexos’s expert (Doc. 194) set out “the following erroneous citations, quotations, or statements of authority” in eleven enumerated items — including twice citing a nonexistent “Hockett v. City of Topeka” and a fabricated quotation attributed to the Federal Circuit’s Microsoft (i4i) decision. Overstock’s reply exposed the errors, and the court ordered the five attorneys of record to show cause why they should not be sanctioned.

What was built

No system was built or deployed here. This is an AI-governance honest-negative: a court’s response to unverified generative-AI output submitted as law. Co-counsel Sandeep Seth admitted he alone used ChatGPT “as a shortcut to find 10th [C]ircuit and Federal Circuit case law” and did not check what it produced. No dedicated legal-AI product is named — the Buether Joe firm even had a policy barring AI for court filings. Senior U.S. District Judge Julie A. Robinson was careful to separate the tool from the conduct: quoting the Tenth Circuit, she wrote that “there is nothing inherently problematic with the use of GenAI in the practice of law,” and that “the violation here is due to the failure to verify that the cases generated by ChatGPT actually exist.” A reasonably competent attorney, she added, “should be aware of the pronounced, well-publicized risks of using unverified generative AI for legal research.”

The outcome

In a 36-page Memorandum and Order filed Feb. 2, 2026 (Doc. 218), the court found that “all five Lexos attorneys violated Rule 11(b) by signing filings that included fabricated and misrepresented legal authority,” and imposed proportionate sanctions: a “$5,000 fine … against Mr. Seth” with his “pro hac vice admission to this Court” revoked, a self-report to state disciplinary authorities, and a certificate of new firm procedures due Feb. 28, 2026; “Mr. Kula is fined $3,000”; “Mr. Joe is fined $3,000” (also a certificate due Feb. 28); “Mr. Cooper is fined $1,000”; and “Mr. Doell is released from any further sanctions” (public admonishment only). The fines are “payable to the Court’s registry within 14 days.” Reuters reported the total as a “combined $12,000” — which the per-attorney figures reconcile to exactly ($5,000 + $3,000 + $3,000 + $1,000). The court noted that attorneys’ fees were not an available sanction because it acted on its own initiative (Rule 11(c)(5)(B)).

Path to green

Every figure is established by the published order and needs no client call. The open items are docket-only and asserted by no claim: whether the fines were paid within 14 days, whether Seth and Joe filed their Feb. 28, 2026 certificates, the outcome of Seth’s disciplinary self-report to Texas and California, and whether summary judgment followed. A green badge — if TIN carries court-adjudicated honest-negatives — would be a human confirming that docket posture; the sanctions and the quotes are already verifiable from the primary.