Twitch started giving Amazon streamer data for AI training

Twitch found itself in the midst of a scandal over new user data processing rules.

The platform has started using streams, VOD recordings, audio, and chat messages to train Amazon's generative AI models. This new feature is enabled by default for all accounts.

The data collection became known after tech support published instructions on how to disable it. Twitch Product Director Mike Minton admitted that if conscious user consent were required, no one would voluntarily enable this feature.

The situation was exacerbated by contradictory statements from Twitch: the platform's official account denied training its own models, while the FAQ directly indicated data transfer to Amazon.

Streamers can disable data collection in privacy settings, but Twitch did not announce this. Currently, during broadcasts, algorithms receive images, streamers' voices, and fragments of games they are showing.