It sounds like a scene from a dystopian future novel - but it's real and it's happening now: The social media giant Meta (i.e. Facebook, Instagram and co.) has started using the publicly shared content of its users to train artificial intelligence from May 27, 2025. Without extra permission. Without payment. And without any real possibility of withdrawal.
Anyone who has not actively objected by May 26 has automatically released their digital self for AI feeding - permanently.
Your posts, photos and likes - soon to be part of the meta-AI
Meta wants to catch up in the race against Google, OpenAI & Co. with its own AI technology. To do this, it needs data. Lots of data. And where do you get it? Quite simply: from its own users.
Everything that has been posted publicly - comments, likes, reels, stories, profile pictures, posts in groups or on pages - can now be used by Meta. And not just for improving the services, but specifically for:
- Training of AI models
- Development of personalized advertising
- Analysis of behavioral patterns
- Reproduction of your writing style or communication behavior
Even deleted accounts are not excluded - once it's in the data pot, it stays there. Get it back? Not possible.
Only those who have actively objected are out - everyone else has lost
What makes this measure particularly controversial is that it was barely communicated and the deadline for objecting was extremely short. Data protection experts are sounding the alarm: the objection was only possible via a hidden form in the login area of the respective platform - and after that it is too late.
Many users were completely unaware of this silent changeover. And now? Access is irrevocably permitted. If you're annoyed, the best you can do is delete your account - but even that won't stop the data being used.
WhatsApp users can (still) breathe a sigh of relief - but for how long?
At least WhatsApp is being left out for the time being. Thanks to end-to-end encryption, private chats are currently not intended for AI training, according to Meta. However, anyone using Meta AI within WhatsApp - for example in new chat functions - should also remain vigilant. Data protection is a promise there, but not a contract.
Welcome to the post-privacy era?
What is happening here is not just another "terms of use trick", but a paradigm shift: content that has been shared carelessly for years is suddenly becoming raw material - for a future business worth billions.
And all this without consent in the traditional sense. No transparency, no access to what happens to the data. No fee, no control. If you don't defend yourself, you lose.
Meta has turned data protection into a one-way street
Anyone who has missed the train is digitally dispossessed - quite legally, quite quietly. And the German constitutional state? It looked the other way. The urgent application against Meta was shot down - with an almost bizarre justification: Meta is economically important.
To be honest, if economic interests are allowed to override data protection, then we have a problem. And a pretty big one at that.
If you want to keep your privacy, you don't just have to be careful - you have to be proactive. Otherwise your next vacation post will become part of the next meta bot. And they may then know more about you than you would like.
