A group of YouTube creators has filed proposed class actions against Amazon, OpenAI, and Apple, alleging that the companies unlawfully scraped millions of copyrighted videos from YouTube to train and commercialize their generative AI video tools, including Nova Reel, Sora, and Apple’s AI models. According to the complaints, the companies used automated systems designed to bypass YouTube’s technological protection measures, including rotating IP addresses and bulk downloading tools, in order to extract large volumes of video content without detection.

The plaintiffs, who include operators of popular channels like h3h3 Productions and Golfholics, argue that this conduct violates both YouTube’s terms of service and the Digital Millennium Copyright Act, which prohibit circumvention of access controls protecting copyrighted works. They further contend that creators relied on these protections when choosing to upload their content, and that the unauthorized use of their videos for AI training deprives them of compensation while enabling companies to profit from their intellectual property. The suits emphasize that once content is absorbed into AI models, it cannot be meaningfully removed, amplifying the alleged harm, and they seek class certification, statutory damages, injunctive relief, and attorneys’ fees. The cases add to a growing wave of lawsuits challenging the legality of how AI companies source and use training data across creative industries.A group of YouTube creators has filed proposed class actions against Amazon, OpenAI, and Apple, alleging that the companies unlawfully scraped millions of copyrighted videos from YouTube to train and commercialize their generative AI video tools, including Nova Reel, Sora, and Apple’s AI models. According to the complaints, the companies used automated systems designed to bypass YouTube’s technological protection measures, including rotating IP addresses and bulk downloading tools, in order to extract large volumes of video content without detection.

The plaintiffs, who include operators of popular channels like h3h3 Productions and Golfholics, argue that this conduct violates both YouTube’s terms of service and the Digital Millennium Copyright Act, which prohibit circumvention of access controls protecting copyrighted works. They further contend that creators relied on these protections when choosing to upload their content, and that the unauthorized use of their videos for AI training deprives them of compensation while enabling companies to profit from their intellectual property. The suits emphasize that once content is absorbed into AI models, it cannot be meaningfully removed, amplifying the alleged harm, and they seek class certification, statutory damages, injunctive relief, and attorneys’ fees. The cases add to a growing wave of lawsuits challenging the legality of how AI companies source and use training data across creative industries.