Amazon might win the right to turn your videos into robot food

Imagine spending 10 years building a channel just to learn your “content” is actually pre-chewed training kibble for an AI that’s about to replace you.

That’s the vibe in YouTubers v. Amazon, where creators are basically asking a horrifying little question: are we humans, or are we datasets with feelings?

Amazon (and the broader AI lab swarm) wants courts to bless the idea that scraping videos to train “video models” is fair use. Translation: “We copied your life’s work, but it’s fine because we did it with a spreadsheet and a straight face.”

Creators are saying the opposite: you don’t get to vacuum up their footage, style, pacing, editing, and voice, then ship a model that can spit out “new” videos in their lane without paying them. Translation: “If you’re building a money-printing machine out of my face, I’d like a cut, not a thank-you note.”

This isn’t about vibes or artistry. It’s about who gets the profit when a machine learns to imitate the exact things that paid your rent.

If the court says training is fair use, the tab gets picked up by creators first, then anyone whose job involves “making stuff” second. If the court says no, AI labs either pay licensing fees, cut deals, or go back to the sewers and pretend they never touched your uploads.

Meanwhile, the “innovation” crowd is doing the usual sermon about progress.

This won’t stay a YouTube thing. If they can call your livelihood “publicly available data,” they’ll do it to your work emails, your code, your voice, your face, your entire annoying little professional existence.