It appears that Facebook has more troubles brewing for it. Now the social media giant has to prepare for a class action lawsuit due to its use of facial recognition technology. According to the lawsuit, Facebook is acquiring biometric information from its users without their permission. This is in reference to the “tag suggestions” feature which tracks the user’s friends in the photos that are uploaded to the platform. The lawsuit argues that this technology is in violence with the USA, Illinois state law, however, the firm said in its statement that this lawsuit has no merit and it will fight it out in the court.
The California judge, i.e., US District Judge James Donato, however, noted that “Facebook seems to believe… statutory damages could amount to billions of dollars.”
The pivotal point, in this case, is the fact that Judge Donato in Monday ruled to certify a class of Facebook users, an integral part of a class action suit. Now, any person belonging to that group is eligible for compensation. The group in question are users in Illinois for whom Facebook created and stored a face template after 7 June 2011,” the court order states. For Facebook, there couldn’t be a worse time to face such a lawsuit. The heat of Facebook chief Mark Zuckerberg being grilled by the US lawmakers about the platform’s collection and usage of user data is yet to simmer down.
For a better understanding of the suit, it was June 2011 when Facebook brought out it’s “tag suggestions” feature which showed people who might be present in the photos that were uploaded by cross-referencing them with an already present database of faces.
Judge Donato’s in the ruling noted that first the algorithm tries to find faces in an uploaded photograph, then the discovered faces are standardized and aligned with size and direction. Then Facebook computes a face signature for every face that it finds, sort of like a mathematical thumbprint, which in this case is the facial structure. Then these face signature is matched with a stored database of user face templates to garner a similar match.