How to opt out of Twitch’s controversial AI training setting

Loading

Twitch has come under fire after creators discovered that a setting allowing their channel content to be used for Amazon generative AI training was enabled by default. Users can disable the setting through Twitch’s Streamer Dashboard, but the opt-out approach has drawn criticism from streamers and viewers who questioned why consent was not required upfront.

The setting applies to a creator’s own channel content, including streams, clips, VODs, images, chats, and other material associated with their channel. However, the controversy has also raised questions about how much content may already have been used and how much control creators have over content involving them on other channels.

Why Twitch’s default AI training setting sparked backlash

Twitch added a setting that allows Amazon to use creator material to train generative AI models. The setting is turned on by default, meaning creators must manually opt out if they do not want their content used for this purpose.

We've added a setting that lets you opt out of having your channel content used to train generative AI content models across Amazon.

Click here to learn more:https://t.co/C3G0Jt6m0D

— Twitch Support (@TwitchSupport) August 12, 2026

The issue gained wider attention on August 12, after Twitch Support posted that users could opt out of having their channel content used to train generative AI models across Amazon. The announcement quickly drew criticism from streamers and viewers, many of whom objected to the feature being enabled by default rather than requiring explicit consent.

The controversy comes during a broader period of tension around artificial intelligence and content creation. Many creators have expressed concern over how AI systems use creative work, voice, video, chat, and other forms of digital content for training purposes.

Some users also questioned what the setting could mean for game developers. Since Twitch streams often include gameplay footage, critics asked whether Amazon’s AI models could also be trained on material from the games being broadcast by streamers.

What is Twitch using this data for?

Twitch chief product officer Mike Minton said during a livestream that users who remain opted in could have any of their channel content used for training. He added that the data collected would not be resold to other companies.

  Twitch's AI learning model has an opt-out option, but it has to be done manually (Image via Elgato)   

Twitch’s own FAQs say that, if users do not opt out, their content could be used to train generative AI models. According to the BBC, Twitch said a person’s audio might be used to refine models that create speech-to-text outputs, which could help improve automatic subtitles on Twitch streams and Amazon videos.

Minton said Twitch was “respecting” users by allowing them to opt out, but many creators and viewers questioned why the setting was enabled by default. He also said he did not know whether VODs had already been used by Amazon for AI model training.

How can streamers opt out of Amazon AI training?

During the same livestream, Mary Kish, Twitch’s head of community, showed viewers how to disable generative AI training through their channel settings.

To opt out, streamers need to go to the Settings tab in their Streamer Dashboard, open Security and Privacy, scroll near the bottom of the page, and toggle off “training for Generative AI.”

  How the opt-out toggle looks for streamers (Image via Twitch)   

According to the BBC, disabling the setting should prevent Amazon from using streams, clips, images, chats, and other channel content to train its generative AI models. However, some users have claimed that the toggle appeared to turn back on after they disabled it, adding further frustration around the rollout.

The backlash highlights a growing challenge for livestreaming platforms: how to balance AI development with creator consent and trust. For Twitch, the issue is not only whether channel content can improve Amazon’s AI systems, but whether creators feel they have meaningful control over how their work, voice, chat, and community interactions are used.