Facebook is all set to make a move that explains how desperate the social media giant is to ‘refine’ and ‘define’ its new identity.
Last week Facebook’s recently rebranded parent company Meta revealed its plans to get rid of its decade-old Face Recognition system. This will led Meta to delete a huge trove of facial recognition templates belonging to more than a billion people as part of a larger initiative to limit the technology’s use across its products.
Meta – earlier known as Facebook – has a change of heart and now believes that the use of the technology in a restricted format would have a larger impact in a positive way. The about-face was described by Menlo Park tech giant as “one the largest shifts of facial recognition usage in technology’s history.”
Users who previously opted in to the setting won’t be recognized in Memories or photos and videos. They will also not see suggested tags with their name in any photos or videos that they might appear in. The company’s Automatic Alt Text tool (AAT), which creates image descriptions for visually impaired users, will no longer include the names of individuals identified in photos.
The discontinuance of Facebook’s program is a result of long-standing privacy and ethical concern about facial recognition. It could be used to target marginalized communities and further racial bias. This led to bans on Facebook’s facial recognition in a number of cities across the U.S., including New Orleans, Boston, San Francisco, and Minneapolis. Amazon declared that it would extend indefinitely a moratorium to law enforcement’s use of its facial recognition systems.
According to the company, the decision was made to balance the benefits of facial recognition with growing social concerns. This is especially important as regulators are yet to establish clear guidelines. Meta stated that it will continue to use face recognition in services that allow people to gain access and verify their identities in financial products, unlock personal devices, or gain access to locked accounts. It also said that could incorporate biometrics in its new metaverse business.
Meta will also be expected to keep DeepFace. This sophisticated algorithm powers Meta’s photo-tagging facial recognition software.
Facebook introduced facial identification in 2010. This allows the company to auto-tag photos and videos on the platforms using names that are based on a template. It notifies users when they appear within multimedia content, and provides recommendations for who to tag.
Although the setting was enabled by default when it launched, the feature was reduced and introduced as an explicit opt-in feature in September 2019. This is where more than a third (or 640 million) of Facebook’s daily active users are reported to have turned on the setting. Facebook reported to have 1.93 billion Daily Active users (DAUs) by the end of Q3 2021.
Meta’s decision not to use facial recognition seems to be an attempt to avoid regulatory scrutiny after years of legal problems, including a lawsuit filed in the U.S state of Illinois against the company. The lawsuit was filed by the Illinois State Bar for violating the Biometric Privacy Act (BIPA). The company is accused to use the tech to identify Illinois residents without their consent. In a class-action lawsuit, the company was ordered to pay $650 million.
This development comes at the time when Facebook is leaving no stone unturned to get rid of its shady past and distance itself from many controversies which have plagued its products over the past years; the rebranding is the most crucial move made in this direction. For example, the company was recently under the lens because it allegedly prioritized engagement and profits over users’ safety and real harms caused by its platforms.