Media & Culture

Twitch lets streamers opt out of Amazon AI training

Twitch finally lets creators block Amazon from using their streams to train AI models

Deep Dive

Twitch has introduced a new setting allowing streamers to opt out of having their content—including posts, streams, and videos—used to train Amazon’s AI models. The update comes after widespread backlash from over 16,000 creators who opposed the practice by default. To disable AI training, users can navigate to Settings > Security and Privacy > Generative AI Training on Twitch’s website or mobile app. However, Twitch clarifies that disabling this option does not prevent the platform from using content for other purposes, such as AI-powered sponsorship tools, recommendations, and AutoMod safety features.

The revelation has reignited debates over data transparency and creator rights, particularly as Amazon’s executives acknowledged that enabling AI training by default was necessary to ensure participation. Mike Minton, Twitch’s head of product, noted that publicly available content is likely used across the industry for AI training, with or without explicit permission. This shift underscores a broader challenge in AI development: the scarcity of high-quality training data. As companies like Meta, Google, and OpenAI increasingly rely on user-generated content for model training, ethical concerns and legal ambiguities continue to mount.

Key Points
  • Twitch now allows streamers to opt out of AI training via Security & Privacy settings, following backlash from 16,000+ creators.
  • Amazon executives admitted enabling AI training by default was necessary to ensure participation, highlighting industry-wide data sourcing challenges.
  • Twitch’s terms still permit content use for other AI-powered features like sponsorship tools and AutoMod, raising transparency concerns.

Why It Matters

This shift forces big tech to confront data ethics and creator rights as AI training demands outpace available high-quality datasets.

📬 Get the top 10 AI stories daily