Picture a tech giant announcing that it intends to utilize your Instagram and Facebook content for training its AI models, without offering anything in return. The company claims you have the option to opt out. However, when you attempt to exercise that choice through their official channels, they simply don't function.
This is the experience being reported by users of Facebook and Instagram. Nate Hake, publisher and founder of Travel Lemming, revealed that he received an email from Meta regarding its plans to use his social media content for AI training, but the link to opt out that was provided was inoperative.
Furthermore, this isn't the only troubling aspect. Meta pledged to respect the wishes of users who do not want their data to be harnessed for AI training. According to Hake, however, when he submitted his request, he received a response stating that the company "cannot take further action" on it.
What is going on?
Back in 2018, during the period when Meta was still known as Facebook, the organization disclosed that it was training AI algorithms using the millions of photos uploaded to Instagram. At that point, Meta AI and the Llama AI models hadn't yet come into existence, but it foreshadowed future issues.
Fast forward to 2025, and Meta is fully invested in AI, a technology that notoriously demands vast amounts of training data. Users on platforms like Instagram and Facebook are now discovering that their personal information is being utilized for AI training, while the geolocation-restricted opt-out tools are failing to work properly.
Meta has a long history of training its models on user-contributed content, but in light of ongoing copyright and privacy concerns, the company has faced significant regulatory scrutiny. In June of last year, it announced plans to halt the use of EU user data for AI training.
However, less than a year later, the company indicated that it would resume collecting user data, including photos and comments, from users located in the EU and the UK to support training for Meta AI. The company justified this move by claiming it is a standard practice adopted globally.
In an official blog post, the company stated, "We’re following the example set by others including Google and OpenAI, both of which have already used data from European users to train their AI models."
Allow them to opt out, if feasible
In mid-April of this year, Meta announced plans to inform users in the EU and the UK about changes to its AI training policy. The notification process was set to include an in-app alert as well as an email notification.
According to Meta, these notifications would feature a link granting users the opportunity to object to their data being utilized in this capacity at any time. The company assured users that the opt-out form would be easy to find and would provide clear information.
Despite these assurances, many users, including Hake, reported that the opt-out link was malfunctioning. He and numerous others took to social media to express their dissatisfaction with Meta's handling of the situation. It remains uncertain whether the issue with the opt-out form is a technical glitch or something more deliberate.
Hake emphasizes the apparent disregard for user choice and criticizes the company's sudden shift in stance. Although Meta recently vowed to respect all objection forms received, including new submissions, Hake’s efforts to express his concerns were met with a claim that no further action would be taken. Meta promotes a transparent approach regarding the use of social content for AI training, but there seems to be a lack of follow-through on these commitments.
