Meta recently launched Meta AI in the European Union, although the company had to settle for an inferior experience compared to the rest of the world. By “launched,” I mean that Meta forced Meta AI on all European users whether they liked it or not. That colorful circular button is now present on Facebook, Messenger, Instagram, and WhatsApp, getting a prime placement.
In some instances, the Meta AI button sits atop the UI, and you might tap it by mistake. I guess this would count as interacting with the AI.
If you thought that wasn’t great, you won’t like what’s coming next. Meta announced plans to start training Meta AI on all the content you generate in its apps. That means public content from apps like Facebook and Instagram, not private chats from WhatsApp and Messenger, which are already encrypted and can’t be accessed for AI training or other purposes.
Thankfully, Meta will let you opt out of training the AI, and you should do it immediately.
Meta laid out its plans in a blog post on Monday, explaining how the AI training will work.
Meta will use public posts and comments from adults in the EU, as well as their interactions with Meta AI, to improve its models.
It’s all starting next week when Europeans in the EU block will start receiving in-app notifications and emails explaining the data that Meta wants to use to train Meta AI and “how this will improve AI at Meta and the overall user experience.”
Data belonging to users under 18 will not be used for training.
These notifications are important because they will give you a way to opt out of training. You’ll have a link where you’ll register your objection:
These notifications will also include a link to a form where people can object to their data being used in this way at any time. We have made this objection form easy to find, read, and use, and we’ll honor all objection forms we have already received, as well as newly submitted ones.
Meta says it needs this data to adapt Meta AI to the European market:
That means everything from dialects and colloquialisms, to hyper-local knowledge and the distinct ways different countries use humor and sarcasm on our products. This is particularly important as AI models become more advanced with multi-modal functionality, which spans text, voice, video, and imagery.
Also, Meta notes that it’s not alone in using public data to train its AI, and the process isn’t unique to the EU market. Meta AI will train on user data worldwide.
If you’re using Meta products outside of the EU and forgot to opt out of having your public data and chats train the AI, it’s time to revisit those settings.
I say all this as a now-longtime European ChatGPT user who disabled AI training as soon as OpenAI improved its privacy features to support opt-outs. I’d take the same stance with other AI models, especially those from companies like Meta, whose sole business has been harvesting user data and turning it into piles of cash.
