A proposed class action lawsuit has been filed against Twitch and its parent company Amazon after a streamer accused them of using content created by countless streamers to train generative AI models without their approval or payment.
Streamer Views Lawsuit as Breach of Unwritten Agreements
Warren Pandiscia, a Twitch streamer with about 1,000 followers, filed his legal complaint last week in the Northern District of California, according to Courthouse News. In the 37-page document, Pandiscia asserts that Twitch and Amazon duplicated large numbers of creator streams without proper licensing. He alleges this violated the implied agreements between creators and the platform and also breached state unfair competition laws.
The filing contends that creators were unaware their videos might be used for AI development, and that Twitch and Amazon only recently introduced a way to opt out. Pandiscia says neither he nor other streamers had any indication their videos were being repurposed in this way and that he promptly opted out once given the opportunity. The complaint maintains that, as early as 2024, Amazon and Twitch were already using streamer footage for AI purposes well before they notified creators or enabled any opt-out.
At this time, both companies have declined to offer public statements regarding the allegations or respond to media inquiries.
Community Pushback Over AI Training With Streamer Content
The situation escalated after Twitch acknowledged on August 12 in a post on X that content from creators—such as live broadcasts, archived clips, chat data, and images—is leveraged for training AI systems at Amazon. Twitch clarified that the opt-out feature does not apply retroactively, so material uploaded prior to opting out may still be involved in AI training processes.
Additionally, the opt-out setting works only for individual channels instead of across a whole user account; if a creator appears on another person’s channel that has not opted out, that content can still be included in the AI’s data set.
The community criticized the move to make AI content training opt-out by default. On forums such as Twitch UserVoice, many streamers urgently called for an opt-in system instead, questioning both the decision to activate the feature by default and the process requiring manual disabling.
Mike Minton, Twitch’s chief product officer, responded to these criticisms by explaining, “if it was opt in, nobody would opt in. So it’s going to be on by default, and almost every content service in the world is on by default.” He recognized that this stance was not favored by users, but emphasized Twitch’s willingness to allow opting out, which he described as uncommon among peer platforms. “This is not a fan favourite,” Minton admitted, “but this is where we are.”
Amazon Expands AI Development Beyond Twitch
Amazon’s initiatives in generative AI reach beyond Twitch. Amazon Game Studios previously started a project involving a generative AI-powered game, but the game’s development ceased following significant layoffs and team closures.
After acquiring Twitch in 2014 for roughly $970 million, Amazon now faces renewed scrutiny regarding its handling of AI technology and user-generated content, as both public and legal opposition mount.
The resolution of Pandiscia’s legal action could shape how online services are permitted to use, profit from, and inform users about the applications of their creations in an era dominated by artificial intelligence, as content creators campaign for increased transparency and agency.