FACEBOOK FACE RECOGNITION. To everyone’s surprise, the social network has decided to end its automatic facial recognition system. All ID patterns on photos and videos will be removed.
It is an announcement to say the least surprising that Facebook made this Tuesday, November 2, 2021: to everyone’s surprise, the social network has indeed decided to abandon automatic facial recognition, this system which makes it possible to identify faces – and therefore of people – in photos and videos. A welcome decision, as this technology was perceived by many users as an invasion of privacy. Certainly it was always possible to deactivate this function in the privacy settings of an account, but the procedure was not automatic and many members were upset to see their names appear in the clear on images posted by friends – sometimes in situations that deserved to be kept private… – without their prior consent. This will no longer be possible and that’s good! It is the end of a system put in place in 2010 and used on billions of images.
In the coming weeks, Facebook will thus remove the identification models used in its automatic easy recognition system. The company says data erasure affects a billion people – which is huge. However, it would appear that all of this information – including tags, or labels – are not completely destroyed. In an interview with Wired, a spokesperson for the group said billions of photos tagged with facial recognition over the past decade would retain those tags. According to the magazine, “clues and cues about a person’s social circle gathered from photos and videos using facial recognition would likely remain intact.”. We will have to wait for a clarification from Facebook to know whether or not the company will erase all face identification information collected through facial recognition from its servers in eleven years.
A strategic decision for the image of the Meta group
As Jérôme Pesenti, vice-president of artificial intelligence for Mark Zuckerberg’s group, explains in a blog post, this decision was made “as part of a company-wide initiative to limit the use of facial recognition” in home products. However, the American group does not completely question its facial recognition tool, which can be useful for “people who need to verify their identity or to prevent fraud and identity theft “. But Jérôme Pesenti specifies that it is necessary to evaluate carefully “The Positive Use Cases of Facial Recognition Versus Growing Concerns in Society”. An awareness in the form of a confession, which marks a profound change in policy.
This announcement is in fact the first important and landmark decision of the Meta group, the new name of the Facebook conglomerate which also includes Instagram, Messenger, WhatsApp and Oculus (see our article). She intervenes at the right time, at a time when Mark Zuckerberg’s group, at the heart of several affairs, seeks to create a new identity for itself by regaining a certain virginity. It must be recognized that the road to redemption is likely to be long, because in addition to the already stale sulphurous files, which are scandalous in the United States, it would seem, according to confidential internal documents, that the company was considering developing a new product to appeal to young people aged six to sixteen, who greatly prefer TikTok and other platforms to the aging social network. A discreetly abandoned project, according to an article published by Ars Technica, probably for fear of a new scandal …