- Twitch has introduced a Generative AI Training option, switched on by default, allowing Amazon to use channel content to train models unless creators opt out.
- Opting out does not fully prevent content use; Twitch says content may still be used for sponsorship assistance, recommendation and discovery tools, and moderation systems such as Twitch AutoMod.
- Twitch has not clearly answered basic questions about whether chat logs are included, when the practice started, how authorship, consent, and attribution work, or whether affiliates, partners, and sublicensees benefit.
- With about 7.3 million monthly streamers and 240 million active users in early 2026, the issue matters beyond Twitch, especially as regulators in the EU and Canada examine opt-out data collection under GDPR-style rules after similar fights with Meta and Google.
Twitch has implemented a default setting that permits Amazon to utilize channel content for AI training, igniting significant backlash from creators. Over 16,000 have voiced objections, particularly after Twitch executive Mike Minton suggested that an opt-in model would lead to minimal participation.1
Users can disable this feature through Settings > Security and Privacy > Generative AI Training, but Twitch warns that opting out does not eliminate all data usage. Content may still be leveraged for sponsorship assistance, discovery tools, and moderation systems like Twitch AutoMod.2
The controversy raises pressing questions about privacy and consent, as Twitch has not clarified whether chat logs are included in the data collection or how authorship and attribution will be managed. This situation is particularly relevant given Twitch's vast user base of approximately 7.3 million monthly streamers and 240 million active users as of early 2026. Regulatory bodies in the EU and Canada are also scrutinizing such opt-out data collection practices under GDPR-like regulations, following similar disputes involving major tech companies like Meta and Google.34
“Twitch has not clarified whether chat logs are included or how authorship and consent work, leaving creators in the dark. With 7.3 million monthly streamers and 240 million users, EU and Canadian regulators are scrutinizing this opt-out model under GDPR-style rules after similar fights with Meta and Google.”
