The comments around the under 16 media ban seem to be missing a salient point.
In order to determine if a user is under 16, you need to determine that for every single user … including you!
This means that your personal data will be harvested in order to determine if you are over 16 or not.
They’ll use machine learning to determine based on your behaviour if you’re under 16, then they’ll ask for proof of age. They are required to have multiple options and none of them are allowed to be a checkbox. Sites are also supposed to take measures against circumvention by VPN whatever that means. You can easily circumvent this by not creating an account, unless you rely on the YouTube subscription feed like I do (though I’ve got a fallback now).
I personally support the idea of restricting social media, but banning accounts is the complete opposite of what’s needed. The algorithm should be what’s banned.
Good luck with using AI. The training dataset will be polluted, and so will the data of individual accounts.
My TV is signed into YouTube as me, but all my kids use it to watch Minecraft videos. Google probably knows my age by now, and all this will get flagged as typical viewing for a millenial.
Thank you for making my point.