When the Arizona Supreme Court handed down its ruling in a pair of arson and DUI cases this month, the announcement wasn’t delivered by a judge or spokesperson, but instead by two AI avatars that are virtually indistinguishable from real people.
Daniel and Victoria are the newest computer-generated faces of the Arizona Supreme Court as part of an effort to innovate how justice is communicated to the public.
“I think it’s just an efficient way for us to get news out,” Communications Director Alberto Rodriguez said. “It’s really an opportunity for us to meet the public where they’re consuming their media.”
Rodriguez is the man behind the machine who helped design the voice and appearance of his virtual co-workers. He says the AI technology has allowed his team to shrink the production time for a video news release from up to six hours down to a few minutes.
Though the process has become more efficient, Rodriguez isn’t worried about becoming obsolete, pointing out that every upload still requires a human touch.
“As far as taking the job of a public information officer, I don’t think that’s an issue because it still takes manpower,” he said. “We still have to work with the bench to make sure that we’re getting accurate information.”
Arizona Supreme Court Chief Justice Ann Timmer assures that every word the public hears from the avatars was written by the justices themselves. She hopes the added attention will help improve public confidence in the court.
“For years, we took it for granted that, ‘Of course you trust the courts, of course you trust judges.’ We’re doing our best and these are hard working people,’” she said. “But if people don’t believe that, it doesn’t matter.”
While Daniel and Victoria might become the most forward-facing examples of AI in the legal system, Timmer points out that the law profession is already using AI every day to assist with legal research, document reviews and data analysis.
Still, the use of AI in the legal system is not without controversy. One plaintiff in New York tried to use an AI attorney to argue his case before he was shut down by the appeals panel. Across the country in California, the state bar faced heavy criticism from the legal community after disclosing that some questions on the state’s February exam were written using AI.
As the use of AI in law continues to grow, so do concerns over instances of AI hallucinations and citations of cases that never existed popping up in legal filings.
Timmer maintains that while these concerns are valid, the Arizona Supreme Court knows where to draw the line.
“This AI, at least that we’re using, is not generative,” Timmer said. If the public is concerned “that we’ll use AI to start substituting for judgment, I don’t think that will ever happen,” she said.
