Twitch Ignores User Preferences: AI Training Enabled by Default

I’ll admit it—I’m genuinely upset. Not just because I use Twitch myself—both as a viewer and occasionally as a creator—but because I can’t stand practices like this. Now it turns out that Twitch is simply using our streams and chats for AI training without really asking us. And without giving us a clear way to opt out. It just feels wrong.
The issue is simple: Tom Verrill, Twitch’s Head of Product, admitted in an interview that they don’t even fully know what’s happening to our content. Whether it’s already being used. Whether it’s being used now. And that’s the problem. We, the users, should have control over our own data—but Twitch doesn’t seem to care.
Default Settings: Who Reads Those Anyway?
Twitch did introduce a setting to opt out of AI training—but let’s be real, who spends hours digging through settings? Who reads every last word in the Terms of Service? Most of us just click “Accept” and move on. And that’s exactly the problem. Twitch is exploiting our apathy to collect our data. And all without giving us a real choice.
Then there’s Verrill’s comment: “No one would consciously choose this if they had the option.” That really says it all. He admits that users probably wouldn’t enable the feature—if they even knew it existed. But that’s the issue: we don’t know. Or we don’t know until it’s too late.
What About Existing Content?
The truly disturbing question is: What about the streams, chats, and videos already on the platform for years? Have they already been used for AI training? Twitch says the feature is new. But Verrill hints that historical data might already be in Amazon’s training datasets. That would be outrageous. We’ve spent years uploading content, engaging with our communities—and now all of it is just being used without our knowledge.
P



rivacy vs. Progress: Who Really Benefits?
I get that AI models need to be trained. But who benefits? Sure, Amazon and the tech elite. But what about us? The creators who’ve built communities? The viewers whose data is being harvested? We’re left with vague assurances like “Trust us.” And that’s not enough.
It hits small streamers especially hard—the ones who rely on Twitch for income. They have no power. They can’t just say, “Stop using my data.” Because if they do, they might lose their livelihoods. And that’s just unfair.
What Can We Do?
The community is furious—and rightfully so. On Twitter and Reddit, users are demanding transparency. Some are now streaming under pseudonyms or distorting their voices to make it harder for AI to scrape clear data. Others are seriously considering leaving Twitch altogether—a tough move when it’s their only source of income.
Experts like Dr. Lena Bauer from the University of Hamburg warn of the consequences. If Twitch gets away with this unpunished, others will follow. Other platforms will think: “Why ask users? They’ll just click through anyway.” And then? Then we’ll have no control over our own data at all.
Hope in the Digital Services Act
Maybe the EU can shed some light here. The Digital Services Act (DSA) could introduce stricter rules. If Twitch violates them, heavy fines could follow. But until then? Until then, we have to ask ourselves: Do we blindly trust Twitch? Or do we finally demand transparency and real control over our data?
For my part, I’m going to dig through my settings. I’ll check if I can find the AI training opt-out. And I’ll think seriously about whether I want to stay on Twitch. Because one thing is clear: We deserve respect. And we deserve to be asked before our data is used for some AI experiment.