Twitch exec says opt-in AI training 'nobody would opt in'
Amazon-owned platform auto-enrolls streamers in generative AI data sharing, sparking backlash after CPO's candid admission on livestream.
Twitch streamers discovered this week that their content is being used to train Amazon AI models by default—and the platform's chief product officer openly acknowledged why the company didn't ask permission first.
During an August 13 livestream, Twitch CPO Mike Minton addressed user complaints about the auto-enrollment policy with unusual candor. "If it was opt-in, nobody would opt in," Minton said. "That's honestly the answer."
The controversy erupted after users found a setting buried in Twitch's privacy menus indicating their streams were being shared for generative AI training. The feature is enabled by default for all users, requiring manual action to disable.
Why it matters
Twitch's admission crystallizes a tension at the heart of AI development: companies need massive datasets to train models, but users increasingly object to having their content used without explicit consent. The episode illustrates how platforms are choosing business imperatives over user preference—and betting that friction and inertia will keep most people enrolled even when opt-out mechanisms exist.
Platform defends industry standard practice
Minton and Twitch head of community Mary Kish defended the approach as standard industry practice during the livestream. Minton noted that AI companies including OpenAI and Anthropic likely already scrape public Twitch streams without any notification or consent mechanism.
Kish suggested internal disagreement between Twitch and parent company Amazon, telling viewers: "The opt-out numbers will help us continue to reiterate that our community is not receptive of AI in any form."
Minton added a distinction: "Twitch has not been training models. I can't speak to what Amazon is training or not training as it relates to any specific usage of content."
How to disable AI training on Twitch
Streamers who want to opt out can do so through these steps:
- Click your profile picture in the top right corner
- Navigate to Settings
- Select Security and Privacy
- Scroll down and uncheck "Training for Generative AI"
One significant limitation: if you chat in another streamer's channel, your messages fall under that streamer's settings, not your own. Opting out only protects content you directly control.
Broader AI training landscape
The Twitch situation reflects wider practices across the AI industry, where companies routinely train models on publicly available content without explicit permission. Google has defended similar practices by comparing AI training to "an art student taking inspiration from walking through a gallery," arguing it constitutes fair use.
These practices have triggered multiple lawsuits, though legal challenges have not slowed the industry's appetite for training data. The difference with Twitch is that the platform made the practice explicit through a settings toggle—then defended why that toggle defaults to "on."
The details were first reported by Paul Tassi at Forbes.
This is an original analysis by the Omega editorial team. Source reporting: AI Watch.
Want systems like this working for your business?
Book a Call

