Imagine being falsely accused of a crime—with a video, audio clip, or document that looks completely real but was never real at all. That future is already here.

In a recent California court case, a judge discovered that a key piece of video evidence submitted by the plaintiffs was a deepfake—an AI-generated fabrication designed to look and sound real.

It nearly passed as legitimate evidence.

According to NBC News, across the country, judges are raising alarms: our courtrooms are not ready for the rise of AI-generated evidence. Deepfakes—realistic but entirely fake videos, audio, images, and documents—are becoming easier to make, harder to detect, and more likely to be weaponized in court.

The consequences are terrifying. Innocent people could be convicted. Victims could be denied justice. Trust in the legal system could erode.

Yet despite these warnings, the U.S. Judicial Conference—the body that sets national evidence rules—recently chose not to adopt proposed reforms that would require attorneys to verify and disclose AI-generated materials. Instead, they said existing standards were “sufficient.” But judges on the front lines disagree.

We, the undersigned, demand that the U.S. Judicial Conference and the Supreme Court immediately:

• Reconsider and adopt updated rules of evidence specifically addressing AI-generated content
• Mandate metadata disclosure and AI verification standards for any digital evidence submitted in court
• Create national training programs for judges on detecting and handling deepfakes
• Accelerate the review and implementation timeline for AI-related rule changes.

The longer we wait, the more vulnerable our courts become.

Truth, justice, and due process depend on reliable evidence. AI deepfakes are a direct threat to that foundation—and if we don’t act now, the damage could be irreversible.

Sign this petition to demand urgent reform before it’s too late. U.S. courts must be equipped to tell real from fake—because lives and liberties depend on it.