Amazon has trained its generative AI models on Twitch content for at least two years, with users automatically enrolled.
Twitch now offers an opt-out option, available in account security settings.
Users' streams, clips, chats, and channel images can be used to improve Amazon's AI systems unless they manually disable it.
What happened
Twitch announced today that users can opt out of allowing Amazon to use their streams, VODs, clips, chats, and channel images to train Amazon's generative AI models. The opt-out is available in account security settings at www.twitch.tv/settings/security. Users are automatically enrolled in AI training by default.
Why it matters
Amazon has been using Twitch content for AI training for at least a couple of years without offering an opt-out, according to Twitch's then-chief monetization officer Mike Minton, who confirmed the practice in 2024. The delayed opt-out gives users control over whether their content contributes to AI systems that could improve captions, speech-to-text, and other content generation across Amazon's services.
What to watch
Users who wish to prevent their content from being used must manually opt out; the default setting keeps training enabled. The support page notes that opting in allows content to be used for "future Gen AI model improvements" and gives an example of audio being used to refine speech-to-text models.
Ask the AI about this article →
Amazon has been quietly using Twitch content to develop generative AI models for more than two years, with users unaware they were enrolled by default. In 2024, when asked directly by The Information, Twitch's then-chief monetization officer Mike Minton acknowledged the practice, saying it occurred "within the bounds of user trust" and privacy regulations. The announcement of an opt-out today represents a shift toward transparency, though the default-enrolled model means users must actively claim control rather than explicitly consent before their content is used. Twitch frames the benefit of allowing training as improvements to features like automated captions, which could help users across Twitch and Amazon's broader ecosystem. The timing of the opt-out—more than two years after the practice was already underway—suggests the change may reflect growing user expectations around AI training disclosures and consent, though the body does not state an explicit regulatory or competitive trigger for the announcement.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Analyst Ming-Chi Kuo says Nvidia has revived the Rubin CPX AI accelerator with a substantially redesigned arch…
A UK study by UK AI Security Institute and Limbic AI surveyed 6,474 British adults
Broadcom's Clayton Donley says companies are doing mission-critical work with AI agents quickly, but without t…
Bank of England governor Andrew Bailey warned that advanced AI poses risks to financial infrastructure in a le…
OpenAI released a new evaluation framework on July 17, 2026, urging companies to measure AI ROI by 'useful out…
As AI agents perform real business tasks, 'Agentic Identity' (giving each AI a unique employee-like ID) and 'D…
