Apple seeks dismissal of YouTube AI training lawsuit

Apple moved to dismiss or, alternatively, strike a lawsuit filed by three YouTube creators in the U.S. District Court for the Northern District of California, setting up a fight over whether publicly available videos can be used to train AI systems without permission. The case, before Judge Richard Seeborg, was filed on April 3, 2026, by Ted Entertainment, Inc., which owns h3h3Productions, H3 Podcast and H3 Podcast Highlights, along with the operators of MrShortGame Golf and Golfholics.

Apple’s filing says the plaintiffs’ claims fail because publicly available YouTube videos are lawfully accessible under the Digital Millennium Copyright Act and YouTube’s Terms of Service. The company is asking the court to reject the complaint outright rather than let the case move deeper into discovery. A hearing on the motion is scheduled for August 6, 2026, in San Francisco.

The complaint alleges that Apple scraped copyrighted YouTube videos to train its AI models and that some of the plaintiffs’ own videos were included in Apple’s training data. Public claims tied to the case have pointed to an alleged dataset containing 173,536 YouTube videos from roughly 48,000 channels, and to Apple research papers that plaintiffs say indicate YouTube material was used in training. Another dataset drawing attention in the dispute is Panda-70M, a 70-million video-caption-pair collection described in a 2024 paper.

The lawsuit lands as Apple expands its generative AI push. Apple introduced Apple Intelligence on June 10, 2024, describing it as a personal intelligence system that puts generative models at the core of iPhone, iPad and Mac. That broader strategy has made the case a test of where companies believe the line falls between openly accessible online material and content creators’ control over how their work is used.

The dispute also reaches beyond YouTube. If Apple persuades the court that publicly available videos can be lawfully gathered for training, the argument could strengthen other AI companies facing similar copyright and consent challenges. If the creators prevail, the case could push platforms and model builders toward licensing deals, tighter dataset rules and new limits on how the economics of online video and other media are set in the AI era.