Twitch and Amazon Hit With Class Action Lawsuit Over AI Training on Streamer Content

A proposed class action lawsuit filed in federal court this week accuses Amazon and Twitch of using streamer broadcasts, past clips, chat logs, and photos to train generative AI models without consent or compensation. The complaint was filed by streamer and plaintiff Warren Pandiscia, who alleges the practice stretches back to 2024, before Twitch ever disclosed it or offered any opt-out.

The lawsuit follows a Twitch post on X on August 12 in which the company confirmed that creator content is being used to train Amazon's AI models, and that the setting is on by default. Chief product officer Mike Minton told Eurogamer that the default-on approach is deliberate. "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," Minton said. "I know this is not a fan favourite, I know it's very upsetting to the community, but this is where we are."

Pandiscia's complaint, first reported by Courthouse News, argues that Twitch and Amazon quietly used creator content for AI training as far back as 2024, before any disclosure or opt-out mechanism was in place. He says he had no reason to suspect his content was being used this way and opted out as soon as the setting existed.

One detail that has frustrated creators is that the opt-out is per channel rather than per account. Streamers who switch off model training on their own channel can still have their content harvested if they appear on someone else's stream that has not opted out. Minton acknowledged the limitation is intentional, telling Eurogamer that the unique setting is "respecting your wishes to opt out of model training."

Neither Twitch nor Amazon has yet responded to Courthouse News' request for comment on the lawsuit. Amazon acquired Twitch in 2014 for roughly $970 million. As reported by Law360, the suit argues that data once used to train a model cannot be "clawed back" through an opt-out, which is a core reason plaintiffs are asking a federal court to step in rather than rely on the setting alone.

The complaint lands at a sensitive moment for Twitch. The platform has spent the last year trying to recover from a post-pandemic dip in viewership, while Amazon has been cutting costs across its wider gaming and entertainment portfolio. A class action over AI training data adds another front to that pressure and gives affected streamers a potential vehicle for damages beyond the platform's own opt-out toggle.

For streamers, the practical reality is that switching off model training does not retroactively remove anything already ingested. Pandiscia's filing argues that opt-out therefore offers no meaningful remedy for content that has already fed the models, which is the central legal hook for turning an individual complaint into a class action. Until Twitch and Amazon respond in court, the per-channel setting is the only available mitigation, and creators who guest on bigger channels will see their work opt them back in by default.