Twitch has added a new privacy control for creators and users who do not want their content used to train Amazon’s generative AI models. The setting gives streamers a direct way to say no to future training of Amazon AI systems that are designed to generate or synthesize media.
What the new Twitch AI setting does
The new control is called “Training for Generative AI.” It lets Twitch users opt out of allowing their channel content to be used for Amazon’s generative AI content models.
When the setting is turned off, Twitch says several kinds of content are covered: “your streams, VODs, clips, stream chats, and pictures and text on your channel.” The opt-out applies to “future training” of an Amazon AI model “whose purpose is to generate or synthesize text, audio, images, or video.”
That scope matters because Twitch content is not just video. A channel can include live broadcasts, saved broadcasts, clipped moments, chat activity, images, and written channel material. The new toggle groups those materials under one generative AI training preference.
Where streamers can find it
The “Training for Generative AI” toggle is in Twitch’s settings under the Security and Privacy tab. The source article notes that the toggle was on when found and that Amazon was asked whether that is the default.
The description attached to the setting explains the tradeoff in plain terms. It says users can allow their channel content to train generative AI content models at Amazon. It also makes clear that turning the setting off is limited to that training use case.
In other words, this is not a blanket opt-out from every Twitch or Amazon use of channel content. It is specifically about generative AI content model training.
What still works after opting out
Twitch says other “AI-supported” features will continue to function even if a user opts out of generative AI training. The examples in the setting description include tools tied to streamer growth and monetization, viewer discovery, and community safety.
The description names real-time sponsorship campaign assistance, recommendations, and AutoMod as examples of AI-supported Twitch features that may still use channel content for other purposes described in the Twitch Privacy Notice.
That distinction is important for creators who want to limit training but still rely on platform features. Based on Twitch’s explanation, turning off the training toggle does not disable captions, safety tools, recommendations, AutoMod, or other AI-supported functions mentioned in the setting description.
How chat is handled
The opt-out is tied to the channel whose content is being governed. Twitch says that if you participate in chat on another person’s stream, “their opt-out preferences govern if that chat can be used for training.”
That means a user’s own preference does not necessarily control every message they type across Twitch. If the chat activity happens inside someone else’s stream, the other channel’s setting determines whether that chat can be included in training.
For streamers, this makes the channel-level choice more consequential. Their setting can affect not only their own broadcasts and channel materials, but also the chat content that appears in their stream environment.
Why this matters for creators
The new Twitch AI opt-out gives streamers a clearer lever over one specific use of their content. Generative AI training can involve many types of media, and Twitch channels often combine all of them in one place: live audio, video, chat, clips, images, and written descriptions.
For creators, the practical question is not whether AI-supported features exist on Twitch. Twitch’s own description says those features can still apply. The key question is whether a creator wants their channel content included in future Amazon generative AI model training.
The answer now lives behind a single switch in Security and Privacy settings. Streamers who want to limit that training use can turn the toggle off, while those who leave it on allow their channel content to be used for Amazon’s generative AI content models.