Courts Set the Terms for AI: October 2026 Industry Roundup

The Texas Supreme Court heard argument in a wrongful termination case over whether a plaintiff can submit an AI-generated transcript of a deposition as evidence. The trial court struck the transcript because a certified shorthand reporter did not prepare it, and the Ninth District Court of Appeals affirmed. Citing Courthouse News Service, the ABA Journal reports that the justices seemed inclined to allow such transcripts, partly because verified video recordings of depositions can help confirm their accuracy. The court's answer will decide whether a transcript prepared without a certified reporter can be used as evidence in Texas.

The Courts and Tribunals Judiciary published a position statement on judicial use of AI, issued by Sir Colin Birss, the Lead Judge for Artificial Intelligence. Its five principles keep decisions with judges: AI may assist but must not displace reasoning, the evaluation of evidence, or judicial discretion, and every AI output used in judicial work must be reviewed, verified, and approved by the responsible judge, who remains fully accountable. The judiciary calls its approach cautious but positive, supported by pilots, evaluation, training, and transparency, and Birss said lessons from those pilots shaped the statement.

A three-judge panel of the Arizona Court of Appeals held that a trial judge should not have allowed an AI-generated video of Christopher Pelkey, an Army veteran killed in 2021, at the sentencing of the man convicted of manslaughter in his death. Pelkey's family created the video, which showed an avatar of him talking about forgiveness and addressing the defendant directly. The panel found that considering it rendered the sentencing fundamentally unfair, upheld the conviction, and ordered a new sentencing hearing.