Streaming platform Twitch has quietly enabled its parent company, Amazon, to train generative AI models using years of creator content, sparking swift backlash across the online broadcasting community. According to reports from TechCrunch and Business Insider, user data including livestreams, video-on-demand recordings, clips, stream chats, and associated channel media have been fed into Amazon’s artificial intelligence infrastructure by default. The practice, which went largely unnoticed by creators for years, has left prominent streamers urging peers to disable the hidden privacy setting immediately.
The controversy broke out when Twitch executives addressed an online audience of nearly 3,000 users during an official livestream. Twitch Chief Product Officer Mike Minton and Head of Community Mary Kish fielded hostile feedback in the chat, where hundreds of viewers posted anti-AI messages. When pressed by users demanding to know why the data collection was set to opt-in rather than opt-out, Minton offered a direct rationale reported by TechCrunch: “If this was opt-in, nobody would opt in. That’s honestly the answer.”
How Twitch Streamers Can Opt Out of Amazon AI Training
For creators looking to block Amazon from utilizing their channels for future model development, Twitch has provided a manual opt-out mechanism buried within account controls. According to TechCrunch and Business Insider reporting, users must navigate to their standard channel settings rather than the creator dashboard, select the security and privacy tab, locate the option labeled “training for generative AI,” and toggle it off. Twitch’s official documentation notes that opting out ensures that future streams, VODs, clips, chat logs, and channel pictures and text will be excluded from upcoming generative training runs for text, audio, image, and video models developed across Amazon.
Despite the implementation of this toggle, executives admitted uncertainty regarding historical data harvesting. When asked on the livestream whether creator archives had already been scraped and processed to train Amazon models, Minton stated, according to Business Insider, “I don’t know what Amazon in this case has done in terms of model training, and what they’ve used and not used. I’m not responsible for Amazon’s training efforts.” Kish added context regarding industry standards, noting that other major platforms like Meta similarly scrape public user content from Facebook and Instagram to train proprietary AI systems.
Community Backlash and the Push for Creator Protections
Prominent figures within the digital broadcasting industry wasted no time condemning the policy shift on social media platforms like X (formerly Twitter). Streamer Legundo posted a widely shared warning urging colleagues to check their configurations, writing, “EVERYONE make sure to turn this off and tell Twitch we don’t want this crap!” Business Insider noted that Amazon, which acquired Twitch in 2014, declined requests for comment as it accelerates its AI infrastructure investments to compete with frontier model developers like OpenAI and Anthropic.
Industry observers point out that the situation reflects the immense value platform owners place on raw human audio and video data. As tech giants race to scale generative systems, user-generated content across live broadcasting remains a prime resource. However, community advocates argue that automatic enrollment undermines creator autonomy, leaving thousands of broadcasters to manually police their privacy settings after the fact to prevent their voices and likenesses from fueling corporate AI growth.
As discussions around data rights continue, creators are advised to check their account privacy settings immediately to ensure their channel archives remain exempt from future data harvesting. Share your thoughts on Twitch’s opt-out policy in the comments below.
Keep reading