In the quiet architecture of digital consent, Twitch has chosen the path of least resistance — and most data. By defaulting all user audio, text, and images into Amazon's AI training pipeline, the platform has made visible a tension that defines this era: the gap between what users create and who ultimately profits from it. The platform's own chief product officer acknowledged that genuine consent would have yielded almost nothing, a candid admission that reframes the policy not as a feature, but as a confession about the limits of voluntary participation in the attention economy.