The photo and video sharing platform Instagram announced Thursday that it is testing new ways to verify the date of birth of its users, in particular thanks to an artificial intelligence tool estimating age via facial recognition.
The social network, which is part of the Meta group (also the parent company of Facebook, WhatsApp and Messenger), has teamed up with Yoti, a British start-up that is developing a facial feature recognition algorithm.
Specifically, minors residing in the United States who use Instagram and wish to change their age to 18 or older will have to justify their approach by filming themselves and sending a video to Meta.
The snippet will then be analyzed by Yoti’s algorithm to determine the age before being deleted.
Yoti says his tool’s margin of error is about 1.5 years for 13-19 year olds. The figures provided by the company show more inaccuracies for female faces and people with dark skin color.
The company guarantees that its technology does not in any way allow the identification of a subject or the recovery of personal information.
Another option explored by Instagram is to ask three of its contacts (all of whom must be of legal age) to certify the age indicated in the application.
Young users already had and still have the possibility of providing an identity document (destroyed within 30 days) to prove their age.
Many underage people bending the rules
The minimum age to create an Instagram account is 13, but many underage people circumvent this legal limit by lying about their date of birth.
Since 2021, the platform has required all of its users to communicate their birthday and asks for confirmation of age to be able to access certain content deemed inappropriate for a very young audience.
It also implemented various parental control tools, including the ability to limit screen time or schedule breaks.
These devices mark an evolution of the discourse of Instagram, which previously believed that it was not responsible for age verification.
His boss, Adam Mosseri, told US lawmakers last year that he thought it made more sense for parents to take care of it. rather than asking every app, and there are millions of them, to verify age.
For some associations for the protection of children’s rights and several politicians, the measures recently put in place are however insufficient.
Instagram had been shaken in 2021 by the revelations of a former Facebook employee, Frances Haugen, who had leaked documents showing that the heads of the network knew the risks of the application for the mental health of minors.
Radio Canada
Source: Radio-Canada
