Bad news for target users (Facebook, Instagram, WhatsApp). After admitting that exploiting all Facebook and Instagram publications since 2007 to train their AI models, Mark Zuckerberg’s company is now interested in the photos he has never shared.
In fact, as Techcrunch reports, several American users from the Internet on Facebook have noticed, trying to publish in their history, messages asking them if they wanted to opt for “cloud treatment.” This option would allow Facebook “to select the media media of its gallery and download them regularly from (SUS) servers.” Facebook can retouch the images of the movie with the AI or offer collages.
Conditions of blurred use
However, by pressing “authorizing”, users accept the conditions of use of goal AI. This allows artificial intelligence to analyze “the media and the facial characteristics” of these unpublished photos, as well as the date on which these photos were taken and the presence of other people or objects in them. Meta also has the right to maintain and use this personal information, without clearly specifying the useful life or the limits of use.
“We explore the means to facilitate the exchange of content for Facebook users testing content suggestions ready to be shared and organized from a person’s film,” says Maria Cubeta, company spokesman.
However, the conditions of use of goal AI, in force since June 2024, do not guarantee that these photos will never be used to train future models. Unlike Google, which is explicitly committed to not using Google Photos personal data for its generative AI, a goal is still deliberately vague.
However, users can disable cloud treatment in Facebook configuration. This will gradually eliminate unpublished photos of servers after 30 days.
Source: BFM TV
