Privacy activists, including the Open Rights Group (ORG) and None of Your Business (Noyb), have slated Meta’s plans. Regulators in the UK put it on hold until recently.
Here’s what you can do to stop Meta’s AI from gobbling up your data — and the details on that viral Instagram image.
What does the viral Instagram image show?
The viral Instagram template spreading false claims about Meta’s AI policy can easily be shared in Stories and Reels.
Meta / Saqib Shah
A growing number of Instagram users have been duped into posting an image they believe will exempt their content from Meta’s AI data harvesting. The picture in question is a dense block of text, styled like a legal notice, but it’s far from the real deal.
The statement in the image reads: “Goodbye Meta AI. Please note an attorney has advised us to put this on, failure to do so may result in legal consequences.
“As Meta is now a public entity, all members must post a similar statement. If you do not post at least once it will be assumed you are okay with them using your information and photos. I do not give Meta or anyone else permission to use any of my personal data, profile information or photos.”
There are a few telltale signs that this is a hoax. Firstly, it looks like a classic example of copypasta, the 2000s-era trend that saw the same, spammy message rapidly disseminated through email and Facebook. It’s also repetitive, contains pseudo-legal babble, and is downright misleading.
The message states that “Meta is now a public entity,” even though it made its Nasdaq debut in 2012 and changed its name to Meta in 2021 (switching its stock ticker from FB to MVRS, for “metaverse,” in the process).
A poorly phrased Instagram sticker won’t prevent Meta from accessing your posts for AI training purposes. Below, we reveal what’s going on and how to stop it.
In a previous email to non-UK users, Meta said it was “getting ready to expand” its AI to more regions, including its new and improved virtual assistant and the other features it announced last year.
As part of the launch, it started using users’ data for a new purpose beyond targeted advertising; namely, to train and improve its artificial intelligence systems. These include the ChatGPT-style large language models it has released over the years, the most recent of which was Llama 3.
Virtually everything you share on Facebook and Instagram is up for grabs, from your posts and photos (with their captions) to the messages you send the AI chatbots once they go live.
However, Meta has promised it will “not use the content of your private messages with friends and family” to power its AI.
Meta recently canned its AI chatbots based on celebrities and fictional characters
Meta
Is Meta allowed to do this?
To enforce the new rules, Meta says it is relying on a legal basis called “legitimate interests”, which essentially gives businesses a lawful way to process user data without the need for explicit consent under the GDPR. That, by the way, is the EU and UK law enforced in 2018 that is designed to bolster online privacy and give individuals more control over their information.
While it isn’t as simple as unchecking a box in your settings, there are ways you can try to avoid having your data hoovered up by Facebook’s AI.
As required by the GDPR (the General Data Protection Regulation is a European Union law that regulates how organisations collect, store, and manage personal data), Meta is allowing users to object to their information being used in this way. To do so, you can fill out a short form on Instagram’s website, which asks for reasons why you want to opt out, with some personal information such as your country of residence and email.
Which? magazine has details about how to find Meta’s objection form here.
Meta says it will “review objection requests in accordance with relevant data protection laws” and will honour successful appeals “going forward”.
Meta began testing more generative AI features for its social networks late last year, including search, ads and business messaging tools
Meta
But, here’s the catch: despite your protests, Meta may still use your data to train its AI in some cases, whether you use its platforms or not. This will affect you if you appear in or are mentioned in posts, photos, and captions shared by Instagram and Facebook users.
What to say when opting out
Targeted advertising is sinister enough to make you feel like you’re being spied on, but how do you take issue with a supposedly benign AI model? Well, it may help to get a better grasp of how and when a company can use legitimate interest before filling out your objection request.
What is legitimate interest?
Broadly speaking, a company must have a clear reasoning for processing your data under the legal grounds of legitimate interest, be it preventing fraud or improving its services. At the same time, businesses must demonstrate that processing data is necessary without a person’s consent. Crucially, they must also balance their reasons against their users’ privacy rights. If the privacy risk is high, they cannot rely on legitimate interest.
All of this must be done with transparency in mind and users must be given the right to object to their data being processed.
Therefore, in this specific case, a Facebook or Instagram user could express their concern about the lack of control over how their data is used in AI training. They could mention the potential for unforeseen consequences or misuse of data in future applications — after all, AI tools have been known to spout misinformation and even hatred in the past.
A person could also argue that Meta’s AI development process lacks transparency. Ask yourself: has the company made it clear exactly how your data is being used or what kind of AI models it contributes to?
It may even be worth suggesting that Meta explore other ways of data collection, such as anonymised data or synthetic data sets for AI training, instead of relying on personal information.
After all, other companies such as Google and Microsoft prioritise publicly available data for AI training over personal information. They also typically require explicit opt-in from users before incorporating their data into training processes.
What if your objection is rejected?
If you’re unhappy with Meta’s response to your objection, you can file a complaint with the Information Commissioner’s Office (ICO), the independent supervisory authority for data protection in the UK. It can investigate Meta’s practices and potentially issue fines or enforcement.
While less common, you could explore legal action against Meta if you believe its use of your data for AI training violates your privacy rights under GDPR. This route is usually for more serious cases and can be expensive, so consulting a lawyer specialising in data privacy is recommended.