Twitch Streamers Sue Amazon Over Secret AI Training Data Use

A proposed class-action lawsuit filed on August 20 in the Northern District of California accuses Twitch and parent company Amazon of covertly harvesting millions of videos, chats, clips, and audio files to train generative AI models without consent. Plaintiff Warren Pandiscia, a Connecticut streamer, seeks damages, restitution, and an injunction, following Twitch's August 12 admission that it uses channel content for AI training and its automatic opt-in default, which Chief Product Officer Mike Minton defended by saying, "If it was opt-in, nobody would opt-in." The lawsuit highlights that the opt-out setting applies only per channel, leaving opted-out streamers vulnerable if they appear in other channels, and that Amazon's $970 million acquisition of Twitch in August 2014 is now under scrutiny.

August 25, 2026 , Inside AI — A Connecticut streamer has filed a proposed class-action lawsuit against Twitch and parent company Amazon , alleging years of covert data harvesting for generative AI training. The complaint landed in the Northern District of California on August 20 . Plaintiff Warren Pandiscia claims Twitch copied millions of videos, chats, clips, and audio files without consent. The lawsuit seeks damages, restitution, and disgorgement of profits. It also asks for injunctive relief to stop further unauthorized use. The legal action follows Twitch's August 12 admission that it uses channel content to train AI models. That same day, the platform updated its Terms of Service and introduced an opt-out setting. Every channel was automatically opted in. During a livestream, Mike Minton , Twitch's Chief Product Officer, defended the default. He said: "If it was opt-in, nobody would opt-in. That's honestly the answer." Mike Minton, Chief Product Officer, Twitch The quote captures the core tension. Streamers see forced participation as a rights grab. Twitch sees opt-out as a practical necessity for AI development. Why the opt-out shield has a fatal gap The lawsuit identifies a structural flaw in Twitch's privacy setting. It applies only per channel. If an opted-out streamer appears in another channel's chat or broadcast, Amazon still collects their data. Pandiscia argues Twitch cannot obtain consent from every participant in captured communications. He also notes creators cannot reclaim two years of historical data already consumed by Amazon's models. The complaint says Amazon had a strong commercial incentive to scrape content. Instead of negotiating licenses, it allegedly took data to improve speech-to-text captions and other AI services. Pandiscia opted out immediately after finding the buried setting. He says the practice cost him money, property, and control over his original work. A decade-old acquisition now faces fresh scrutiny Amazon bought Twitch for roughly $970 million in August 2014 . The acquisition gave Amazon a massive video library and real-time interaction data. That asset is now central to the legal fight. The complaint references 2024 statements by Minton, then Chief Monetization Officer. He admitted Twitch used content for AI "prototyping" and claimed it stayed within privacy rules. Streamers say they were never directly notified. The lawsuit accuses Twitch and Amazon of unjust enrichment, unfair business practices, and contract breaches. It represents all Twitch video streamers whose content was used without permission. Legal experts note class certification could be difficult. Proving damages across millions of creators with varied content types is complex. But the case may pressure platforms to rethink default opt-in policies for AI training. Twitch has not commented on the lawsuit. The platform's opt-out tool remains available, but critics say it fails to protect creators appearing in third-party channels.