The debate over how major companies train their AI models may seem like a thing of the past, from those first “wild” years (2023–2025) when the first popular generative artificial intelligence models emerged, along with all the uncertainty about how platforms were using content generated by their users and how those users could defend their right to privacy.
We have already seen it in cases involving Meta and OpenAI.
But nothing could be further from the truth. It is now 2026, and Amazon has just reignited this debate. It has done so through its streaming platform, used every day by millions of users around the world.
And of course, it has sparked controversy.
Twitch users have been greeted with a message from the Amazon-owned platform explaining that a new settings option has recently been added, allowing them to opt out of having their channel content used to train generative AI content models at Amazon.
In other words, if you do not disable that option, which is checked by default, your channel content will be used to train those AI models.
Specifically, the button you are supposed to disable includes the following text:
“Allow your channel content to be used to train generative AI content models at Amazon. Turning off this setting does not prevent Twitch and Amazon from using your channel content for other purposes described in the Twitch Privacy Notice, including AI-assisted Twitch features that benefit the community by supporting creator growth and monetization (such as real-time assistance for sponsorship campaigns), viewer discovery (such as recommendations), and community safety (such as AutoMod).”
Yes, you read that correctly. Turning off that option will not prevent your content from being used to train AI. It will simply prevent it from being used to train generative AI.
In fact, in its FAQ section, the company explains in a bit more detail what this means:
“By allowing training, your content may be used to improve future generative AI models. One example of what happens when you allow your content to be used to train a generative AI content model is that your audio could help refine models with speech-to-text capabilities, improving captions on Twitch as well as on Amazon.”.
Wonderful. Especially for Amazon, of course.
Meanwhile, if you disable that option, your streams, clips, stream chats, images, and text from your channel will not be used in the future training of a model developed by Amazon whose purpose is to generate or synthesize text, audio, images, or video.
Adding even more fuel to the controversy, during a livestream intended to answer questions on the matter, Twitch executives did not clarify what would happen to content created before this option appeared, nor did they explain how long Amazon has been using data from Twitch broadcasts for this purpose.
At this point, if you are already eager to rush to your Twitch channel and turn that option off, the process is fairly simple: go to your channel settings, select the Security and Privacy tab, scroll down, look for the “training for generative AI” option, and disable it.
Almost exactly one year ago, the focus of a similar controversy was Google, through its video platform YouTube. Thanks to a CNBC leak, we learned that it had been using its massive library, made up of 20 billion videos, to train Gemini, VEO3, and other AI models. In that case, Google explained that it was only using a subset of those videos and that it was doing so legally, in compliance with agreements with creators and media companies.
“We have always used YouTube content to improve our products, and that has not changed with the arrival of AI. We also recognize the need for safeguards, which is why we have invested in strong protections that allow creators to protect their image and likeness in the AI era, something we remain committed to doing,” a YouTube spokesperson said at the time.
Image: ChatGPT
Your email address will not be published. Required fields are marked *
Δ