After Twitch revealed it opts users into AI training by default, streamers consider boycotts and tag broadcasts to protest. Amazon faces reputational risk as creators decry 'data grab.'
Amazon's livestreaming platform Twitch triggered a wave of hostility this week after revealing it had enabled a new policy by default: any video broadcast on Twitch could be used to train generative AI models, unless users manually opt out.
The news broke Monday when the official Twitch Support account on X posted about a new dashboard option that lets users block their content from Amazon's AI training. The post carried two implicit announcements – Amazon intends to feed Twitch data into its generative AI models, and that the option is enabled for every account by default.
According to Twitch's FAQ, the data collected may train a model "whose purpose is to generate or synthesize text, audio, images, or video."
Several users quickly took to social media to report that the opt-out setting kept switching itself back on after they turned it off.
Twitch's head of community, Mary Kish, and chief product officer, Mike Minton, held a live stream Tuesday to address the backlash. Minton was unusually candid.
"There's an honest answer, and I think most of you can probably appreciate this," Minton said, responding to viewers who demanded the settings be opt-in rather than opt-out. "If it was opt-in, nobody would opt in."
Kish acknowledged the community's anger and said similar practices are standard across the livestreaming industry.
The Policy
The default opt-out policy covers all content streamed on Twitch – not just the video but also the audio, chat logs, and any on-screen text. Amazon has not specified which of its generative AI products the data will feed into, though the company has been investing heavily in AI across its cloud and consumer services.
Creators who want to block their content must navigate to the Security and Privacy section of the Twitch dashboard, locate the "Training for Generative AI" toggle, and turn it off. Some streamers complained the toggle reset after they saved their settings.
The Fallout
Seattle-based broadcaster Lance Icarus expressed concern about feeding his playthroughs of indie games that are proud to not use generative AI into Amazon's models. He also worried about streaming with voice actors who fought for rights against AI training.
Another Seattle streamer, Will Overgard, called the approach a calculated move to extract data.
Kish noted during the stream that Twitch and Amazon are watching the numbers to see how many broadcasters actively opt out. At time of writing, streamers have begun tagging their broadcasts with "AIOptedOut" or "NoAI" to signal their stance. Some audiences have pledged to stop watching channels that do not opt out.
The controversy risks further erosion of Twitch's already declining market share. According to a July report from StreamsCharts, YouTube Live and TikTok Live have both overtaken Twitch's audience share, though Twitch still handles roughly half of livestreamed video game content.
For Amazon, the reputational damage could compound existing concerns among creators about data use, content moderation, and platform monetization. Any regulatory push on generative AI training data disclosure would add another layer of risk.
Discussions continue about the creator community's next steps. Some streamers have called for a coordinated opt-out week. Others are evaluating whether to leave the platform entirely.
Drafted by a large language model from the source reporting linked above, then screened by automated publishing checks. It is not read by a journalist before publication. Some articles cite our Alpha Score. Verify prices and figures against the original source. Educational coverage, not personalized advice.