Twitch 實況主現在可以選擇退出 AI 訓練計畫
Twitch streamers can now opt out of AI training programs
Updated at: August 13, 2026 at 12:45 AM
2026年8月12日，Twitch引入了一項新的隱私設置，允許實況主阻止他們的內容被用於訓練亞馬遜的生成式人工智慧模型。
On August 12, 2026, Twitch introduced a new privacy setting that allows streamers to stop their content from being used to train Amazon’s generative AI models.
然而，此功能默認為開啟，這意味著每位創作者都會被自動加入。
However, this feature is enabled by default, meaning every creator is automatically opted in.
此項更改涵蓋了實況直播、影片以及聊天訊息，在社群中引發了巨大的反彈。
This change covers livestreams, videos, and chat messages, sparking significant backlash from the community.
Twitch認為，AI訓練對於現代平台的發展至關重要，其產品長Mike Minton表示，如果系統採取「選擇加入」(opt-in)制，參與度將會過低。
Twitch argues that AI training is vital for modern platform development, with Chief Product Officer Mike Minton stating that if the system were 'opt-in,' participation would be too low.
此外，即使實況主選擇退出，如果其他創作者未關閉該功能，他們在該用戶直播間的訊息仍可能被收集。
Furthermore, even if a streamer opts out, their messages in another user's stream might still be collected if that creator has not disabled the feature.
這場爭議凸顯了關注數據所有權的創作者與旨在透過用戶生成內容推動生成式AI能力的平台之間，日益增長的摩擦與矛盾。
This controversy highlights the growing friction between creators concerned about data ownership and platforms aiming to advance generative AI capabilities through user-generated content.
