Twitch users are pushing back against Amazon after learning the platform has been using their streams, chats, and other content to train AI models by default. The company announced an opt-out option on Wednesday, but the move has sparked anger over why the data collection was ever switched on without consent.
The setting, which allows Amazon to use creator and audience data to train its artificial intelligence systems, is turned on by default for every user. Twitch's FAQs state that if users do not opt out, their content could be used to train generative AI models. Generative AI creates new content such as text, images, and video, and powers chatbots like OpenAI's ChatGPT and Google's Gemini.
Why the default-on choice angered users
Twitch's Chief Product Officer, Mike Minton, defended the approach during a stream, but his reasoning drew sharp criticism. He said the company chose default-on because of a simple calculation: "users by letting them opt out but on why data was collected by default, he admitted:" that few would voluntarily sign up. Minton added, "If it's opt-in, nobody would opt-in. That's the honest answer."
That admission did little to calm the backlash. One streamer wrote, "On by default is criminal.... the AI narrative push is so draining," capturing the mood of many in the community. Another viewer said, "Nobody would opt in because nobody wants to feed AI with our creativity and content." The stream's chat was flooded with hundreds of critical comments, with users questioning why Twitch would treat their creative work as free training material.
Some also raised concerns for game developers. Because the feature trains AI on video games played by streamers, developers' intellectual property could end up in Amazon's models without their consent. Twitch said a person's audio might be used to refine models that create speech to text, which could improve automatic subtitles on Twitch streams and Amazon videos. But that potential benefit did not outweigh the unease.
How to turn off the AI training setting
For those who want out, the process is straightforward. Users go to Settings in the Streamer Dashboard, click Security and Privacy, scroll down, and toggle off "training for Generative AI." Opting out prevents Amazon from using streams, clips, images, chats, and other channel content for AI training.
Mary Kish, Twitch's Head of Community, provided a tutorial on how to disable the feature. But some users reported problems. One person said, "Went in and toggled the switch off, exited, went back in, switch is still on," suggesting the setting may not stick. Twitch has not yet addressed those reports.
Kish acknowledged the negative reaction during the stream. She said, "We don't expect you to be happy or excited about this. I don't expect anyone to react to this favourably." She also noted that data collection for AI training has become an industry standard, a point that did little to reassure critics.
Stay ahead of the AI curve
The most important updates, news, and content — delivered weekly.
No spam. Unsubscribe anytime.
Unanswered questions about data already collected
It is not clear when Amazon began collecting Twitch users' data for AI training. Minton said he did not know if users' data had already been scraped for training. He was unsure what Amazon "has done in terms of model training and what they've used and not used." That uncertainty has fueled further distrust among creators.
Minton did say that data collected will not be re-sold to other companies. But that promise did not address the core concern: that Amazon, which owns Twitch, is using the platform's content to build its AI products. Amazon bought Twitch for nearly $1bn (£740m) in 2014, and since then the streaming service has grown into a major hub for live content. Amazon runs a host of AI services and has heavily invested in AI development to compete with Google, Meta, and others.
A broader industry pattern
Twitch is not alone in this approach. Kish pointed out that data collection for AI training has become standard practice across the tech industry. Google, Meta, and other giants have all faced scrutiny over how they gather data for their models. But the default-on nature of Twitch's setting, combined with the personal and creative nature of streaming content, has made this case particularly contentious.
One viewer said Twitch had an "opportunity to set an industry standard" to push back against data collection for AI. Instead, the company chose the opposite path, and the community is making its displeasure known. The article was published on 13 August 2026, and updated 26 minutes later as more users reported issues with the opt-out toggle.
What happens next
Twitch has not said whether it will change the default setting or investigate the reports of the toggle turning back on. For now, users must manually opt out and hope the setting holds. The broader question of whether Amazon has already used their content remains unanswered, and Minton's uncertainty on that point has done little to reassure anyone.
The backlash shows no sign of fading. Streamers and viewers alike are watching closely to see whether Twitch will adjust its approach. The company's response so far, a default-on setting with an opt-out buried in the dashboard, has only deepened the distrust.