Twitch has announced that its users can now prevent Amazon from utilizing their creative content to train generative artificial intelligence models. This update comes after confirmation that the tech giant, which acquired the streaming platform in 2014, had been leveraging streams, clips, images, and chat logs to fuel its AI ambitions. While Amazon views these datasets as essential tools to compete with rivals like Google and Meta, the revelation sparked immediate backlash from a community built on independent creativity.
The new system operates on an opt out basis, meaning all users are automatically enrolled in the data collection program unless they manually change their settings. During a recent broadcast explaining the policy, Twitch Chief Product Officer Mike Minton offered a candid reason for this approach, admitting that if the feature were opt in, virtually nobody would sign up. He emphasized that while the company intends to respect those who choose to leave the program, the default setting remains active across the board.
Community leaders at Twitch acknowledged that the news would not be well received by their audience. Mary Kish, the head of community, walked viewers through a tutorial on how to disable the tracking while noting that such data collection has essentially become an industry standard. She admitted bluntly that she did not expect users to be happy or excited about the practice, a sentiment echoed by thousands of commenters who argued that streamers should not have to fight to protect their own intellectual property from being fed into an algorithm.
Despite these assurances moving forward, significant questions remain regarding what has already happened behind the scenes. When pressed during the live session about whether user data had already been harvested for existing models, Minton conceded that he was unsure exactly what Amazon had previously utilized for training purposes. For many creators who have spent years building massive followings on the site, this lack of transparency adds another layer of frustration to an already contentious debate over digital ownership in the age of AI.
