The legal profession has shown mixed feelings about AI, even as its use by lawyers and courts has surged. The debate between early adopters and skeptical old-schoolers was on full display in a single footnote in a February 2026 election law decision by the U.S. Court of Appeals for the Fifth Circuit. 

The footnote, authored by Circuit Judge Edith H. Jones, took aim at U.S. District Court Judge Xavier Rodriguez of San Antonio, questioning whether he had inappropriately used AI in making his rulings. Here is footnote 1: 

There are several things worth noting about this footnote, beginning with its target. Rodriguez has been a federal judge since 2003, and before that a Justice on the Texas Supreme Court and a partner at Fulbright & Jaworski. He has for years been a sought-after voice on technology in the courts. When the Wall Street Journal profiled how judges are using AI, he was a natural subject, and candid about it.  “We need to be heading into the future,”  he told the Journal.  The article also noted that he “still always makes the final decision in cases himself.” 

It is not surprising, then, that the Journal interviewed him. What is surprising is that the footnote does not reflect what the article actually said. 

The article described how Rodriguez uses AI to “summarize legal filings, decipher the facts, and write questions to lob at lawyers during hearings.” It also described how he “first used AI experimentally to see how it would have adjudicated a seven-week trial he had recently completed”.  The Journal article described how Judge Rodriguez described his use of AI: 

“In deciding the case, involving Texas’ election law, he and his law clerks had sifted through hundreds of thousands of exhibits, reviewed testimony from more than 70 witnesses and researched relevant case law. The entire process took 10 months before the judge issued a 140-page opinion. Out of curiosity, he ran that same evidence through an AI tool. Within minutes the computer had produced a first draft that included the findings of fact that had taken his team weeks to produce. While the final result wasn’t perfect and needed some fact-checking, Rodriguez has been using AI ever since.”

With this in mind, a few things about Jones’ footnote deserve attention. 

First, Jones acknowledged the connection was unconfirmed. “Whether it was this case is uncertain” appears in her own footnote. Yet she still attached that insinuation to her opinion. 

Second, the WSJ article did not support the insinuation regardless. Rodriguez’s experiment was conducted after his opinion was already written. He ran the same evidence through AI to compare its output against his own finished analysis. The article made that sequence clear. 

Rodriguez has been exactly the kind of judicial voice the profession needs on AI adoption: rigorous, transparent, and willing to do the comparative work that responsible experimentation requires. 

Fifth Circuit — La Union del Pueblo Entero v. Abbott — Feb. 12, 2026 by JetStream AI Advisory