Texas Supreme Court: Weighs Admissibility of AI Deposition Transcripts

• The Texas Supreme Court heard arguments on allowing AI-generated deposition transcripts as evidence.
• A lower court struck an AI transcript in a wrongful termination case because it lacked certified shorthand reporter preparation.
• The employee's attorney argued certified video recordings sufficiently ensure accuracy for AI transcripts.
• The employer's attorney raised concerns about increased discovery costs and security risks with third-party AI platforms.
• Justices appeared open to permitting AI transcripts, especially if verified video recordings are available as an "escape hatch."

Texas lawyers should closely monitor the Supreme Court's decision regarding AI-generated deposition transcripts, as a ruling allowing them could significantly impact discovery costs, evidentiary challenges, and the required methods for verifying deposition accuracy in future litigation.

The Texas Supreme Court recently convened to hear arguments concerning the admissibility of AI-generated deposition transcripts as evidence, a case that could significantly shape future Texas AI deposition rules. The specific dispute originates from a wrongful termination claim filed by a Montgomery County man against his former employer. During the discovery phase, the employee utilized an AI platform to transcribe a deposition of the employer's representative.

However, this AI-generated transcript became a point of contention when the Montgomery County trial court struck it from the record in April 2025. The court's decision was based on the fact that the transcript had not been prepared by a certified shorthand reporter. This ruling was subsequently upheld by the Texas Court of Appeals for the Ninth District in an unsigned denial of the employee's appeal. The appellate court clarified that their concern was not with the content of the deposition itself or its use as evidence, but solely with the uncertified, AI-generated nature of the written transcript.

Notably, the employer did not raise an objection to the use of the transcription software at the time the deposition was conducted. Their challenge to the AI-generated transcripts evidence only emerged several weeks later, when they requested the trial court to strike it. This delay prompted Chief Justice Jimmy Blacklock to question the necessity of the case reaching the high court, remarking that either paying for a certified reporter or simply allowing the AI transcript would have been a more cost-effective solution for the parties involved.

During the Supreme Court's deliberations, justices probed the attorneys on both sides regarding the implications of allowing AI in witness depositions. Bradley Snead, representing the employee from Houston-based Wright Close Barger & Guzman, contended that the AI-generated transcript's accuracy was sufficiently guaranteed by the accompanying certified video recording of the deposition. He distinguished this scenario from other concerns about AI "hallucinations" in the legal system, such as the fabrication of case citations, asserting that such issues were not relevant to the present matter. Snead argued that under existing rules, the use of such transcripts should be permissible unless the court opts to establish new regulations.

Conversely, Stephanie L. Bryant of Houston-based The Bays Firm, counsel for the employer, urged the court to consider the broader ramifications of accepting AI-based transcripts. She warned that such a precedent could lead to an exponential increase in discovery costs and introduce substantial security and confidentiality risks, particularly when third-party AI transcription platforms are involved. Bryant maintained that while in-house AI software might be acceptable, using external services like the employee did would violate current Texas rules of procedure due to these concerns.

Justice Jane Bland pressed Snead on the absence of human attestation for the AI-generated product, questioning how reliability could be assured without a human standing behind the transcribed material. Meanwhile, Justice J. Brett Busby challenged Bryant's assertion about increased costs for defendants and reduced costs for plaintiffs, suggesting that existing proportionality rules might adequately address such financial imbalances. Justice Kyle Hawkins echoed the sentiment of the appeals court, proposing that the video recording could serve as an "escape hatch" to verify or correct any potential errors in the AI transcript, thereby mitigating concerns about its accuracy.

The Texas Supreme Court's eventual decision on the admissibility of AI-generated transcripts evidence carries significant weight for legal practice across the state. The justices appeared largely receptive to the idea of allowing these transcripts, especially if they are accompanied by verified video recordings that can serve as a definitive reference point for accuracy. This inclination suggests a potential shift towards more flexible Texas AI deposition rules, acknowledging technological advancements while attempting to safeguard evidentiary integrity.

Should the court rule in favor of allowing AI-generated transcripts, Texas lawyers should closely monitor the decision. Such a ruling could profoundly impact discovery costs, potentially lowering them for some parties while introducing new challenges related to evidentiary standards and the methods required for verifying deposition accuracy in future litigation. The debate highlights a growing tension between leveraging cost-saving technologies like AI in witness depositions and maintaining the stringent standards of legal procedure and evidence. The outcome will set a precedent for how AI in the legal field, particularly for transcription, is regulated and accepted within the Texas judicial system.

Texas lawyers should closely monitor the Supreme Court's decision regarding AI-generated deposition transcripts, as a ruling allowing them could significantly impact discovery costs, evidentiary challenges, and the required methods for verifying deposition accuracy in future litigation.

Source: Original reporting via Courthouse News Service.