A Texas federal judge sanctioned an attorney for AI-fabricated citations in a TCPA case
In McCormick v. Texakoma Financial, Inc., decided June 11, 2026, the U.S. District Court for the Eastern District of Texas sanctioned attorney Amy L.B. Ginsburg after her summary-judgment response cited a nonexistent case, fabricated quotations, and misstated legal principles that appeared to come from generative AI. Judge Amos Mazzant struck the response, imposed a $5,000 penalty jointly on Ginsburg and her firm, ordered CLE and a review of her 2026 filings, and required a verification certification on future filings citing authority.
Records by entity: McCormick v Texakoma Financial
The court found her explanation shifted blame to a law clerk and paralegal, calling it a contrived attempt to avoid responsibility.
Key facts
- What
- In McCormick v.
- Incident date
- Jun 11, 2026
- Who
- McCormick v. Texakoma Financial (plaintiff's counsel)
- Failure mode
- Hallucination
- AI surface
- Copilot
- Severity
- Medium
What happened
In a Telephone Consumer Protection Act case, the plaintiff's response to a motion for summary judgment cited a case that did not exist, included fabricated quotations, and misstated legal propositions in a pattern the court tied to generative AI. Judge Amos Mazzant ordered attorney Amy L.B. Ginsburg to appear with copies of the cases she cited; her explanation attributed the draft to a departed law clerk and a paralegal who allegedly filed the wrong version, which the court found lacked credibility and shifted blame. The court struck the response, imposed a $5,000 penalty on Ginsburg and her firm jointly, ordered her to verify the authorities in every 2026 filing she signed, required CLE on the ethical use of AI, and mandated a signed verification certification on future filings, with the order published in the Federal Supplement.
What broke inside the model
- 01 · TriggerA user asks for a fact, a citation, or a figure.
- 02 · Model stepThe model writes a fluent, confident answer.
- 03 · Control gapNothing ties the claim back to a real source.
- 04 · FailureA fabricated fact ships as if it were verified.
- 05 · ConsequenceThe false claim reaches a customer, a court, or the public.
Confidence holds, and even spikes, as the claim detaches from any source.
The model generated fluent, authoritative-looking legal citations and quotations with no basis in real case law, the archetypal hallucination failure. Presented in a legal brief and filed without independent verification, fabricated authorities passed as genuine because the output read as competent and specific. The error surfaced only when opposing counsel and the court checked the citations.
What it cost
Sources
- PressTexas Court Orders Sanctions After AI-Generated False Citations Surface in TCPA Litigationreceivablesinfo.com
- PressHOLY SANCTIONS!: GenAI Sanctions Come to the TCPAWorldnatlawreview.com
Cite this entry
https://failureindex.ai/failures/mccormick-texakoma-attorney-sanctioned-ai-citationsAI Failure Index. "A Texas federal judge sanctioned an attorney for AI-fabricated citations in a TCPA case" (FI-0713). Realm Labs. https://failureindex.ai/failures/mccormick-texakoma-attorney-sanctioned-ai-citations (indexed Jul 10, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0713. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm would have caught this
- Prism
- OmniGuard
- AI Detection & Response (AIDR)
Realm flags the moment a model commits to a fabricated citation or quotation by reading the hallucination signature in its output, and can hold the passage for verification before it is relied upon. In a drafting workflow, that check surfaces the invented authority at generation time rather than at the point a judge does.