Professor Rebecca Wexler recently testified before the U.S. Senate Judiciary Committee’s Subcommittee on Criminal Justice and Counterterrorism at a hearing on the use of artificial intelligence in criminal investigations and prosecutions.
“Although AI tools present exciting opportunities to render the legal system more accurate and equitable in some respects, they also, in their current form and use, present troubling obstacles to fair and open proceedings,” she told senators.
To help fix the problem, Wexler said, Congress should do two things: Require AI tools be available for auditing by independent researchers and also stop both prosecutors and defense attorney from invoking the trade secret privilege to block access to relevant evidence. That evidence should be required to be disclosed, she added, under reasonable conditions to protect the interests of the company that produced the technology.
Wexler, who has written extensively about data, technology, and secrecy in the criminal legal system, is a faculty co-director of the Berkeley Center for Law & Technology and was a senior policy advisor at the White House Office of Science and Technology Policy during the spring of 2023.
In her testimony, Wexler gave several examples of tech-enabled evidence with problems. At least six people, she said, have allegedly been wrongfully arrested or jailed after being identified by AI-based facial recognition software. And UC Berkeley researchers Hany Farid and Sarah Barrington recently found that an AI-trained system for estimating height and weight from a photo of a person yielded worse results than work done by a person with no special training.
“Allowing A.I. to enter criminal courtrooms without sufficient scrutiny is dangerous,” Wexler said. “While A.I. has the potential to improve law enforcement efficacy, it can also cause harmful errors with high-risk high stakes consequences for life and liberty.”