But What Kind Of Data Are They Collecting?
It’s not just me saying this. On Meta’s website, anyone can read:
“As we shared in September, we use publicly available online and licensed information to train AI at Meta, as well as the information that people have shared publicly on Meta’s products and services. This information includes things like public posts or public photos and their captions. In the future, we may also use the information people share when interacting with our generative AI features, like Meta AI, or with a business, to develop and improve our AI products.”
Now you tell me if the title is clickbait or just an unsettling reality. Well, at least they cross their hearts about not reading our private messages.
But hey, that’s the bare minimum, right?
Meta privacy measures and user control
Meta claims a commitment to privacy and has established a robust internal Privacy Review process.
This includes evaluating potential privacy risks and implementing measures to mitigate them, they say.
They operate under five core values: privacy and security, fairness and inclusion, robustness and safety, transparency and control, and governance and accountability.
Do you trust them?
What Can Meta Do With Your Data?
Meta uses the information you share on its platforms to power their generative AI models.
These models, which include tools like Meta AI and AI Creative Tools, are designed to enhance user experience. Their goal is to solve complex problems, provide real-time answers, and allow for creative expression.
What does this mean?
The data collected is used to teach AI models to understand relationships between different types of content.
For instance, text models are trained to predict language patterns and generate conversational responses. Image models learn to create new images based on descriptive text inputs.
What’s wild is, these models don’t just learn in some sterile, one-time process. Every time people post something clever, creative, or even a bit odd on a Meta platform, those examples might get tossed into the giant cauldron of training data. It raises a question I’ve been mulling over: is there a line between inspiration and appropriation when machines are the ones “learning” from us?
I always wonder what this means for the originality of a post or photo you thought was just for your friends or family. There’s almost an unspoken trade of culture and ideas, invisibly happening every second. Some people might argue it’s worth it for fancier AI features, while others would rather those digital impressions stay personal. Depends on your threshold for being data-mined, I guess.
Is Everyone Subjected To Meta’s AI training?
Actually, not everyone is subject to having their info repurposed for AI training.
If you live in the UK, any of the European Union (EU) member states, or the European Economic Area (EEA) you benefit from stricter data protection laws.
Why? Businesses outside these regions but processing data of UK or EU citizens must comply with the UK’s Data Protection Act (DPA) and the EU’s General Data Protection Regulation (GDPR).
So, if you live in the UK or EU, initially, you could follow a few steps in the “Settings and Privacy” menu to opt out of Meta’s AI training.
Now, you don’t even need to worry about it. Meta has completely paused their AI model training in Europe, due to the tightening of European data regulations.
Is this a step back for European Innovation?
According to Meta, it is. Meta admitted to being disappointed and defined the situation as “[…] a step backward for European innovation, competition in AI development further delays bringing the benefits of AI to people in Europe.”
The CNIL disagrees! The French data protection authority published its first recommendations on applying the GDPR to AI system development. According to CNIL, it’s all about innovating while respecting people’s rights—what a novel concept, right?
They’re confident that with the right guidelines, we can have our cake and eat it too. We can enjoy the perks of AI without trampling on privacy.
What If I Don’t Live In The UK Or EU But Don’t Want Meta To Use My Data?
Some people have been posting on Meta’s social media, saying they don’t allow their data to be used for AI training.
Something like: “Hey Meta, I’m not giving you permission to do this. Stop.” And obviously, things don’t work like that.
As I said, we accepted the terms (even if we didn’t read them), and there’s no way to opt-out.
But there’s always “the road less traveled by” (channeling Robert Frost) — you could delete your published content or your entire account and live like the Incas.
Another option is to set everything to private, since the terms only apply to public information. I like this one better.
However, although you can delete content or make accounts private, it’s challenging to erase your digital footprint, especially when others might share information about you.
A little something you can try
Meta has a form hidden in its privacy center called “Data Subject Rights for Third-Party Information Used for AI at Meta.” You can fill it out to ask Meta to delete any personal info used for their AI.
But, here’s the kicker—Meta doesn’t just automatically do it. They review the requests to see if they’re required by local laws to remove your data.
Some folks, like Alanna Smith, a small business owner in Toronto, have tried this route. She had to prove her data was actually used by Meta’s AI, which is kind of like proving your cat isn’t the real boss of the house—nearly impossible!
Smith never got a confirmation, so she’s left wondering if her request even went through. And when asked, Meta’s spokesperson didn’t give any details about how many people have successfully opted out.
So, good luck with that!
Is There a True Way Out?
It depends on what you consider a way out. But, hey, have you heard of the Cara app?
The Cara app is a new social networking platform designed specifically for artists to share their work without the fear of it being used to train AI models.
Cara positions itself as a safe haven for creatives. It ensures that images posted on its platform are not scraped for AI training purposes.
Cara achieves this by implementing features like “NoAI” tags on all posts, signaling to AI scrapers not to use the content.
While these tags might not be foolproof against determined data miners, they represent a significant step towards protecting artists’ work.
Should You Be Worried About Your Data?
Alright, let’s get real here. Should you be worried about Meta using your data to train their AI models?
Despite Meta’s core values and privacy measures, it’s totally understandable to have some concerns about your data being scooped up for AI training.
Sure, Meta offers transparency and control features, but let’s be honest—the sheer amount of data they collect can feel a bit overwhelming.
The real worry boils down to the balance between control and transparency versus the vast amount of data Meta collects and uses.
On one hand, Meta’s generative AI models promise some seriously cool features and better user experiences. But on the other hand, this comes at the cost of extensive personal data usage.
And for those of us outside the protected regions, opting out of data usage for AI training is usually not an option.
So, what do you think? Is the trade-off worth it for those nifty AI features, or does the idea of your data being used make you want to rethink your social media habits?
