Amazon will train on Twitch streamers’ content by default, unless they opt out

Twitch has officially confirmed that it will begin utilizing creator content to train generative AI models for its parent company, Amazon. The announcement has triggered a wave of intense criticism from the platform’s user base, primarily because the policy is set to "opt-out" by default, meaning streamers are automatically enrolled in the data-harvesting program unless they manually change their settings.

For Amazon, the vast library of Twitch content represents a goldmine of training data, offering thousands of hours of video and audio that can be used to refine AI capabilities. However, for the creators who spend hours each week broadcasting their likeness and voices, the move feels like a violation of their digital autonomy.

The "Opt-In" Controversy

During a recent broadcast on the official Twitch channel, Head of Community Mary Kish and Chief Product Officer Mike Minton faced a barrage of questions from nearly 3,000 viewers. The chat was flooded with anti-AI sentiment, with many users questioning why the company chose an opt-out model rather than requiring explicit consent.

Minton provided a candid—if controversial—explanation for the company’s strategy:

"Why is it not opt-in? That’s what everybody is spamming in chat. I get it. ‘Let me opt in versus making me opt out.’ Well, there’s an honest answer… If this was opt-in, nobody would opt in. That’s honestly the answer."

This admission highlights the fundamental tension between platform owners and the creators who drive their traffic. Twitch’s community is largely skeptical of generative AI, viewing it as a technology built on the unauthorized scraping of creative works, including books, images, and videos.

Navigating the Confusion

Twitch’s initial communication regarding this change was perceived by many as opaque. Rather than explicitly stating that Amazon would be training on user data, the company framed the update as the introduction of a new setting that allows users to "opt out" of having their channel content used for Amazon’s generative AI models.

This phrasing left many streamers wondering if their past broadcasts had already been ingested into Amazon’s systems. When pressed on the matter, Minton admitted to a lack of clarity regarding the company's historical data usage:

• The Uncertainty: Minton stated he could not confirm whether past content had already been used for training, noting, "I don’t actually know the answer to that question because I don’t know what Amazon […] has done in terms of model training."

Industry Context and How to Protect Your Content

Twitch is certainly not the only tech giant leveraging user data for AI development. Meta has faced similar scrutiny for using public posts from Facebook and Instagram to train its AI models. While users in the U.K. have some recourse to opt out, those in other regions often find that the only way to prevent their data from being used is to make their accounts private—a move that is often impossible for professional creators who rely on public visibility to monetize their work.

Kish defended the inclusion of the opt-out toggle as a direct response to the community's feedback, stating that it serves as "a reflection of reacting to this community’s voice that you are not wanting to train Gen AI models."

How to Opt Out

If you are a streamer looking to prevent your content from being used for Amazon’s AI training, you must take action manually. Follow these steps:

1. Navigate to your Channel Settings (ensure you are in your settings, not the Creator Dashboard). 2. Locate the Security and Privacy tab. 3. Scroll down to the section labeled "Training for Generative AI." 4. Toggle the setting off.

As the industry continues to grapple with the ethics of AI training, this incident serves as a stark reminder of the ongoing power struggle between platforms seeking to innovate with AI and the creators who provide the essential data that makes that innovation possible.