Arizona Supreme Court Leverages AI Avatars to Boost Public Accessibility
PHOENIX, (AP) — In a pioneering move, the Arizona Supreme Court has introduced AI-generated avatars to communicate court rulings, marking what is believed to be the first instance of a state court using artificial intelligence to connect more effectively with the public.
The court is deploying two avatars, Daniel and Victoria, to disseminate news about every decision made by the justices. The move is rooted in a broader effort to foster public trust and confidence in the judicial system, according to court officials. The court’s Chief Justice, Ann Timmer, has made public trust a central element of her agenda, recognizing the need to engage with the public more actively through digital platforms.
“We serve the public better by saying, OK, we’ve issued this decision,” Timmer said, “Now, let us help you understand what it is.”
Timmer told The Associated Press that the court is always seeking to improve the way it shares information.
Created with Creatify software, the AI avatars bring the court’s news releases to life. Videos featuring Daniel or Victoria are now published with every ruling, with the potential for use in community programs, civics education, and Access to Justice projects in the future. The court began distributing releases in October, summarizing and clarifying rulings.
Court spokesperson Alberto Rodriguez noted that the shift to AI-generated videos has significantly accelerated the information dissemination process. While producing a video typically requires several hours, an AI-generated video is ready in about 30 minutes. The court is considering adding more AI reporters in the future, Rodriguez said in a news release.
The justice who authors a legal opinion also drafts a news release, and the entire bench must approve the wording. Subsequently, the justice collaborates with the court’s communications team to develop a script for the avatars. Rodriguez clarified that the avatars do not interpret original court decisions or opinions.
Daniel and Victoria’s names and appearances are intended to represent a broad demographic, Rodriguez said. Their purpose is not to appear as real people, and the court explicitly acknowledges their AI origins through disclaimers. The court is actively exploring a range of emotional tones, speech cadences, and pronunciation styles for the avatars, as well as Spanish translation options.
Mason Kortz, a clinical instructor at Harvard Law School’s Cyberlaw Clinic, described the court’s new cyber employees as “quite realistic.” He also suggested that the video disclaimers should be prominently featured to ensure that their AI nature is not missed.
Asheley Landrum, an associate professor at Arizona State University’s Walter Cronkite School of Journalism and Mass Communication, expressed that although the avatars currently feel robotic, a format that more closely resembles real dialogue and storytelling might be more engaging.
Copyright 2025 by The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.
