Case D-48 · AI Sanctions Case Tracker

Ma v. Cogdell Law Firm PLLC

Texas District Court, Travis County, 419th Judicial District, No. D-1-GN-25-001604, order signed October 22, 2025 (Judge Maya Guerra Gamble; attorney Lawrence Chang)

The case

Ma v. Cogdell Law Firm PLLC, Tex. Dist. Ct., Travis County, 419th Judicial District, No. D-1-GN-25-001604, Order Granting Defendants' Motion for Sanctions signed October 22, 2025 (Judge Maya Guerra Gamble). Primary order: https://aboutblaw.com/bjVR ; reported: https://news.bloomberglaw.com/litigation/texas-attorney-sanctioned-over-ai-generated-citations-in-filing

What happened

Plaintiff's counsel Lawrence Chang (Choate & Associates PC) filed a motion for reconsideration containing AI-generated citations that did not exist or had been altered. Chang first blamed "over-reliance" on a Westlaw research tool, then realized he had pasted his draft into ChatGPT's legal tool to improve his writing: ChatGPT changed his citations and added references that were not there originally. He admitted in a September 23 filing that he used AI and did not check the veracity of the citations it produced. The underlying case: Chang represents an acupuncture business alleging legal malpractice by attorney Dan Cogdell over a $2.3 million settlement with the federal government.

The rule and the sanction

Judge Guerra Gamble granted opposing counsel's sanctions motion, finding "good cause supports the imposition of sanctions in order to deter such conduct in the future, which amounts to serious misconduct that threatens to impugn the integrity of the courts and the administration of justice." Sanction: $2,000 in attorneys' fees and expenses payable to opposing counsel (Rusty Hardin & Associates LLP); Chang must write a letter outlining where his AI use erred and how he will avoid doing so in the future; and the reconsideration motion containing the bogus citations was struck in its entirety.

How it could have been avoided

Never paste a near-final draft into a chatbot "to improve the writing": LLMs rewrite citations as freely as prose, and previously-correct Westlaw citations can come back fabricated. Re-verify every citation after any AI touch on the document, even if the tool only "edited" your prose. And get the story straight early: the initial (incorrect) blame on Westlaw undermined Chang's credibility before the court.

---

Sources