Twitch is being sued for allegedly using streamer content to train Amazon's AI. I read the lawsuit. And the timeline is WILD: In 2024, Twitch's CPO acknowledged that Twitch content was already being used by Amazon for AI development And then August 2026, we all know what happened Twitch rolled out a new setting allowing creators to opt out of having their content used to train Amazon's generative AI You're opted in by default Twitch also updated its Privacy Notice to explicitly say information may be used to develop or deploy generative AI models 8 days later, streamer Warren Pandiscia files a 37-page proposed class action against Twitch and Amazon (I also went down a rabbit hole trying to figure out who Warren Pandiscia is and how he ended up as the plaintiff. There's an interesting story there, but several identity connections aren't solid enough for me to publish yet) So here's what we know. According to the lawsuit, Warren has: - 900+ followers - Streamed 1,000+ hours - Been on Twitch for nearly 10 years - Invested tens of thousands of dollars into creating content His argument is basically "creators made this content for Twitch" "They never agreed to have years of streams, VODs, clips, audio, images and chat turned into training data for Amazon's AI" Twitch has a pretty strong defense. The ToS gives Twitch a very broad license to content, including rights to reproduce, modify, sublicense and create derivative works But here's where it gets interesting: If those old Terms already gave Amazon permission to train AI on Twitch creators... - Why introduce an AI opt-out now? - Why explicitly add generative AI to the Privacy Notice now? - And what exactly was Amazon doing with Twitch content between 2024 and 2026? That's what I want to know... Because this isn't just about Twitch. SAG-AFTRA spent years fighting over consent, compensation and control when performances are used for AI Now we're starting to see the same fight come to the creator economy Except streamers don't have a union negotiating for them... And every major platform is sitting on years of human-created video, audio, text and conversations that could be incredibly valuable training data This lawsuit could help determine how much of that creators already signed away. This is massively conequential for all platforms...