The live streaming platform Twitch is facing a class-action lawsuit over its use of streamers' live streams, clips, chat logs, and photos to train Amazon, its parent company's generative AI models, without consent or compensation. Warren Pandiscia, a streamer with about a thousand followers, filed a 37-page complaint last week in the Northern District Court of California, accusing the platform of violating the creator's implied agreement and state unfair competition laws.

"Default On" Sparks Community Outrage

The controversy centers on the platform's opt-out mechanism, which is effectively useless. Twitch recently announced that it would use generative AI within the platform, with the feature turned on by default and requiring users to manually go into settings to turn it off. Chief Product Officer Minton bluntly responded, "If it was voluntary, no one would join, so we have it on by default." What worries streamers even more is that the opt-out option is not retroactive and is tied to individual channels rather than accounts—meaning even if you opt out, your content can still be used for training if you appear in a live stream from another channel that hasn't opted out.

The complaint states that the two companies had already secretly used streamers' content as early as 2024, long before any disclosure or opt-out mechanism was introduced. Pandiscia said he opted out as soon as the setting was available and emphasized that he had no reason to suspect his content was being used this way beforehand. This lawsuit highlights the core conflict between the creator economy and AI training: when user agreements include "default consent," individuals' control over their own content is being quietly eroded by platforms.