Judges Discuss AI Challenges in Evidence

Federal judges gathered in Boston have been alerted by tech industry insiders about the rising difficulty in verifying AI-generated images and videos. This conversation comes as a judiciary committee is contemplating whether new rules are needed for evidence that may be deepfakes rather than authentic material.

The U.S. Judicial Conference’s Advisory Committee on Evidence Rules has, since 2023, been investigating how courts should approach evidence that might be altered or fabricated using artificial intelligence. During a recent meeting, representatives from OpenAI and Meta expressed concerns that detecting these fake materials is becoming increasingly challenging.

“AI images are only getting better, and I do not think that there will be a scientific way going forward to detect them,” shared Shira Anderson, policy counsel at OpenAI.

Meta’s director of product management, Devin Norder, reinforced this notion, stating that the newest AI models create media so realistic that distinguishing them from actual content is nearly impossible. “Media generated from frontier models, at least in the last few generations, are indistinguishable by humans from authentic or AI-generated media,” Norder commented. He also noted that creating convincing fakes is not only simple but also inexpensive: “And the cost and the time it takes to create this media really are quite trivial at this point.”

Both companies have developed tools aimed at verifying the authenticity of visual content, and some lawmakers have begun taking similar steps. Legislation in California and the European Union mandates “watermarking” to identify AI-generated work. However, Anderson warned that no universal system can effectively capture every instance, as different verification tools created by various companies don’t interact with each other. “My guess is that this absolutely will continue to get better. I do not think it’ll ever be bulletproof,” she concluded.

Interestingly, despite these concerns, the committee hasn’t encountered significant evidence of deepfakes entering courtrooms. A survey from the Federal Judiciary Center earlier this year revealed that among 931 judges, merely 15 had faced a case where evidence was disputed as a deepfake. Jesse Furman, a U.S. District Judge from Manhattan and chair of the committee, noted this disconnect between what people anticipate and actual occurrences. “There’s a lot of academic commentary and other commentary predicting that courts will be dealing with deepfakes, but we haven’t actually seen a whole lot of that yet,” he remarked.

This discrepancy has fueled debate among committee members regarding the necessity of an AI-specific evidence rule. Judge Valerie Caproni from Manhattan argued that existing laws suffice and that there is no need to reinvent the wheel for AI: “We’ve dealt with bad actors presenting fake evidence since Cain and Abel. That’s a longstanding problem,” Caproni stated.

Regular Americans are indeed affected by AI deepfakes, especially concerning fraud schemes. Wynton Hall delves into these risks in his bestseller Code Red: The Left, the Right, China, and the Race to Control AI.

During a Newsmax interview, Hall elaborated, “They can now mine your social media and then they build a detailed dossier on your life, and then they use the AI to mimic a loved one so convincingly that your instincts will tell you that it’s real.” He also mentioned that these tools are becoming widely accessible: “These tools are now becoming democratized so that they’re cheap, they’re easy to use, and that means they’re accessible at a mass scale.”

According to the FBI, AI-related scams resulted in over 22,000 complaints and nearly $893 million in losses in 2025.