This week, Facebook began including a message in its Norwegian users' notifications, which was related to artificial intelligence (AI).
If you click on that message, you'll receive a message that Facebook is rolling out its AI tools in Norway.
The message also states that they will use your information to further develop Meta's AI services.
Training AI on lots of data
In order to develop good AI services, the more information the AI is trained on, the better. Therefore, it has long been believed that Meta would have a significant competitive advantage with the content of billions of users.
The challenge is that this is the kind of use of your photos and posts that Facebook never asked for. But instead of asking permission to expand their use, they use something called “Legitimate interest”/“Justified interest”.
This term is a legal designation that comes from the European Union's Privacy Regulations (GDPR). The regulations are complex, but they boil down to the fact that companies can use personal data as long as they have clear consent from the user. Legitimate interest is the exception to explicit consent.
Or as the Norwegian Data Protection Authority describes: “A company may process personal data if this is necessary to protect a legitimate interest that outweighs consideration of the individual’s privacy.”
But can Facebook have access to use photos, videos and other content without paying for them, to build a commercial service, which is in the interests of users?
There is a reservation option. You must then submit a form including justification as to why you do not want your data used. Nettavisen tested this and quickly received confirmation that the request had been accepted and that it would honor the request.
Norwegian Data Protection Authority: – We are concerned
And at the Norwegian Data Protection Authority, they are not convinced that what Facebook is doing now is okay.
-We are concerned about this. Meta will train its AI on users' posts and photos, i.e. content that can be of a very private nature. Perhaps the most natural thing would be to ask for approval before they do it, says department manager Tobias Godin to Nettavisen.
– When Meta chooses not to ask for consent, she takes a legal risk. It is doubtful that they can use “legitimate interest” in this way. Meanwhile, Meta is known for putting users' privacy at stake if there is a chance to increase profits.
In the United States, there has been a great deal of uproar about companies resorting to using information to which they do not have rights in order to build new services.
– Currently, there is a race between technology companies to be the best and largest in the field of generative artificial intelligence. Perhaps the strategy is to get us addicted to their new AI services until the important questions disappear. In practice, very few people will go to protest. Maybe that's exactly what Meta is hoping for, Godin says.
Mita believes they are doing everything right
Nettavisen brought Meta criticism from the Norwegian Data Protection Authority, questioning how much user interest there is in the company being able to build new commercial services based on users' content without compensation.
The company does not respond to this, but points to one press release They posted earlier this week.
In the letter, they only state that they have a legitimate interest in processing the data to create these services.
Viskal does this in a responsible manner that complies with privacy rules, the company wrote.
This approach is consistent with how other technology companies are developing and improving their AI experiences in Europe.
“Web specialist. Lifelong zombie maven. Coffee ninja. Hipster-friendly analyst.”
