The idea that any company can simply say they want to train AI on user data and just do it is insane. But companies have been changing the Terms of Service fully knowing people don’t read or understand (or can do anything about it, including opting-out) for years now.
I’ve said it before, just because a user posted it doesn’t mean they have the rights to it. Meta deciding to train their AI on Instagram user data isn’t just reckless, it actively violates copyright.
Because users can upload whatever they want, they do. And at no point in the upload process must a user acknowledge they own everything about what they post. Therefore, Meta doesn’t have permission from the actual copyright owners of anything. They only have implicit permission from users, who—blindly—agreed to Terms of Service updates that they have no alternatives for anyway.
Think about it. Imagine an account that posts before-and-afters of company logos. That account’s content is fair game? Even if the copyright owners of the logos they’re comparing didn’t have a say? How does that even pass the sniff test?
And companies like Meta will get away with it for several years. It’ll be held up by legal teams and courts forever.
Even though it’s obviously unethical to simply flip a switch and say you have permission to do something with users’ data when the users didn’t know in advance or have the ability to opt-out, they’ll get away with it. Because they’ve been testing the limits for over a decade.
Same goes for Apple, Google, and anyone else playing this game. It’s absolutely morally wrong, and I think there are also valid legal arguments to be made too.
At best, it’s monopolistic behavior that the FTC should be restricting and fining exorbitantly for. It’s something they can only get away with because of their scale. When users don’t have reasonable alternatives in the market and/or ways to migrate data from one service to the other, these companies will continue to take advantage of that.

Ver todos los episodios