A Connecticut-based Twitch streamer has filed a class-action lawsuit against the platform over its recent announcement that creators’ content will automatically be used to train Amazon‘s AI products.
In his suit, Warren Pandiscia argues that “[b]ecause Amazon AI products are commercialized, Amazon had an overwhelming incentive to acquire training data on an unprecedented scale. Rather than negotiate for lawful licenses or seek permission, defendants accessed the Twitch streams and videos to utilize them as a massive dataset necessary to fuel Amazon’s AI products.”
Pandiscia’s claim stems from Twitch’s Aug. 12 announcement that it will opt every single user’s account into consent for Amazon to use their streams, chat comments, etc, to feed AI models. (Not how consent works, btw.) Users who do not actually consent have to go in and opt out of the setting.
Twitch decided to make the setting opt-out instead of opt-in because, as CPO Mike Minton said in a “jaw-dropping” admission, “If it was opt-in, nobody would opt in.”
Despite knowing that, and despite the immense backlash to the policy and Minton’s comments, Twitch has kept the auto-opt-in, and will continue to scrape users’ content unless they opt out.
Pandiscia says this opt-out system isn’t good enough. “By design, defendants never obtain—and their systems are incapable of obtaining—the consent of all parties to the communications they capture,” his suit claims.
But his class-action also addresses a widespread fear among streamers: that Twitch just now added an opt-out to something it’s been doing for years–and even if creators opt out, they’ve already had their work stolen.
Pandiscia alleges Twitch and Amazon have been scraping creators’–and viewers’–content as far back as 2024, without notifying them. That allegation finds ground, considering that in 2024, Minton said Amazon was using content from Twitch “in a prototyping, not in any kind of production scale, capacity.”
Neither company has commented on this accusation. During the aforementioned recent livestream, Minton gave a vague answer, telling viewers, “Twitch has not been training models. I can’t speak to what Amazon is training or not training as it relates to any specific usage of content.”
“Content creators such as plaintiff and the class members will never be able to claw back the intellectual property unlawfully copied and used by defendants to train Amazon’s generative AI,” Pandiscia’s suit says.
He claims breach of implied contract, unjust enrichment, breach of express contract, and unfair business practices, and is asking the court to award injunctive relief plus damages, restitution, and disgorgement of profits.
However this suit shakes out, it could end up being a landmark case with regard to major tech-owned digital platforms and creators’ content. Twitch is not the only platform that has scraped or otherwise interfered with creators’ content for its AI-slop-loving parent company, but gen AI is so new that we don’t have precedents yet to address the posting on a UGC platform → AI scrape → copyright law protections? pipeline. This could help set a precedent.
Pandiscia’s complaint was filed Aug. 20 in the Northern District of California.
Stay up-to-date with the latest and breaking creator and online video news delivered right to your inbox.
Newsletter
Trending Stories
- YouTube is reportedly offering creators millions if they don’t sign deals with Netflix
- Blog Action Day: ‘Big Ideas’ from GOOD and Babelgum
- Moral video creator Dhar Mann responds to the protest launched by some of his regular actors
- After becoming China’s answer to YouTube, BiliBili has its eyes on the rest of the world
- YouTube’s now counting public views from the “first frame,” hoping inflated numbers will help creators negotiate brand deals
