“We will remove individual face recognition templates for more than a billion people,” promises Jerome Bisenti. He heads the Artificial Intelligence (AI) division at Meta (formerly Facebook). The organization gradually descended into a deep crisis – so deep that the entire group was renamed. Civil rights activists have long criticized Meta Face Recognition. Photos and videos uploaded without the express consent of those affected were evaluated.
If you don’t want to be scanned, but still want to use Metas products, you’ll have to look in the settings to turn off face recognition. The legality of this practice has been controversial since the inception of automatic facial recognition on Facebook in 2010. In Europe, Facebook has been recognizing faces since mid-2011.
That same year, the then Hamburg data protection officer, Johannes Kaspar, opened a case against Facebook. But only a year later, the case against Facebook’s facial recognition was suspended. Because the data protection authority in Ireland was primarily responsible.
Ridiculous negotiation with students
Austrian data protection activist Max Schrams, then a law student, contacted Irish authorities. In a move that seemed absurd, the authorities then ordered negotiations between Facebook and the Viennese student.
At the time, Facebook claimed that users would agree to “everything” when they first registered, including having their face biometrics analyzed. The data company has never been able to definitively explain how the third party relates to what can be seen on the uploaded photos. This year, Facebook had to pay each user in the US state of Illinois $350 in compensation for unwanted facial recognition – a total of $650 million.
Facebook originally borrowed the necessary algorithms from the Israeli startup Face.com, before Facebook bought the facial recognition technology without further ado, including the startup. The automatic facial recognition system will be deactivated across the group “in the next few weeks”.
Because after more than a decade, lawmakers and data protection authorities are slowly starting to move. Meta, who had just thought he was recovering from the Cambridge Analytica scandal, was plunged into a deeper crisis as a result of the new discoveries. A whistleblower says Facebook is not compatible with democracy.
Meta requires tagging manually
Meta needs good news. It now comes after 1,300 Nicaraguan government propaganda accounts were deleted on Facebook shortly before the elections Announcing the deletion Facial recognition features.
Pesenti said on Twitter: “This is part of a company-wide restriction on the use of facial recognition in all of our products. It was not an easy decision because there are many places where facial recognition has been very useful to people on Facebook. But we had to weigh that against There is growing public concern about facial recognition as a whole. Regulators are also still working on clear rules for the technology. So we are limiting its use to a narrow area.”
Meta does not completely give up face recognition. Meta would like to continue assessing faces with biometrics in order to unlock devices, and to combat fraud and identity theft or identification in order to gain access to a blocked account. With respect to other areas of application, open discussion is warranted, which he is looking forward to, says Pesenti.
Disabling facial recognition for blind and visually impaired Facebook users has a drawback. Automatically generated alt texts for photos can no longer identify the photo of the person being photographed. At the same time, there is no automatic notification to Facebook users if they can be seen in a photo that someone else has uploaded. Meta wants to encourage users to manually enter who can be seen in photos and videos. Meta, of course, is still interested in collecting data.
(s)
