Twitch is under fire after it was revealed that the platform has been harvesting user content to train Amazon’s artificial intelligence models by default.
In a display of corporate arrogance, Twitch Chief Product Officer Mike Minton admitted during a livestream that the company made the data collection opt-out rather than opt-in because, in his words, 'If it's opt-in, nobody would opt-in.' This admission confirms that the platform knowingly bypassed user consent to feed Amazon’s AI ambitions.
The data being scraped includes streams, clips, images, and chat logs, which Amazon intends to use for its own generative AI services. While Twitch claims this will help with features like speech-to-text, users are rightfully concerned about their intellectual property being used to build products for a trillion-dollar tech giant.
Adding to the frustration, users have reported that the opt-out toggle in the settings menu is unreliable, with some claiming the feature re-enables itself after they have turned it off.
Twitch’s head of community, Mary Kish, essentially told users to deal with it, stating that data collection for AI training is now an 'industry standard' and admitting she did not expect anyone to react favorably.
Amazon, which purchased Twitch for nearly $1 billion in 2014, is desperate to keep pace with competitors like Google and Meta in the AI arms race, and it appears they are more than willing to treat their own creators as raw material to get there.
Tags
