Twitch sparks creator backlash with default opt-out for Amazon AI training
Streamers must now manually disable a new setting to prevent their content from being used to train generative AI models.
Twitch has launched a new privacy setting that allows streamers to prevent their content from being used to train Amazon's generative AI models. The feature is enabled by default for all users, meaning creators are automatically opted in unless they manually disable the setting.
The new toggle covers a wide array of creator data, including live streams, videos on demand (VODs), clips, and channel chat. To disable the feature, users must navigate to Settings > Security and Privacy and locate the "Training for Generative AI" option. However, Twitch noted that opting out does not disable all artificial intelligence on the platform; the company will continue to use data for mandatory platform functions, such as captions and AutoMod.
The Industry Push for Data
This move comes as generative AI developers aggressively seek massive datasets to refine their models. Twitch's approach mirrors a broader industry trend seen across platforms like Meta and YouTube, where companies have faced criticism for utilizing user-generated content without explicit, prior consent. By implementing an opt-out rather than an opt-in system, Twitch prioritizes the rapid acquisition of training data over individual creator permission.
Tension Over Creator Consent
The decision has ignited significant backlash from the streaming community, who argue that the use of intellectual property for AI training should require an active choice. The friction is highlighted by the company's own justification for the default setting. Mike Minton, Twitch's Chief Product Officer, explicitly stated that an opt-in system was not viable, saying, "If it was opt in, nobody would opt in. That's honestly the answer."
This admission underscores a growing adversarial relationship between AI developers and the creative class. For many streamers, the default enrollment feels like an appropriation of their work to build tools that could eventually compete with human creators.
What to Watch
The controversy highlights an ongoing struggle over data ownership and the ethics of AI training in the digital age. As creators push back against "opt-out" defaults, the industry may see an increase in legal challenges regarding intellectual property rights. Whether this policy leads to a migration of talent to competing platforms remains to be seen, but the tension between platform owners and the creators who provide their value continues to escalate.