Amidst the privacy issues that are surrounding Facebook and the growing social backlash against facial recognition technology, a researcher has found that the tech giant is currently testing a facial recognition verification feature using selfies taken by the users.
Hong Kong-based software engineer, Jane Manchun Wong (@wongmjane), posted on her Twitter account that Silicon Valley-based tech superpower, Facebook, is testing a feature that uses selfies taken by the users while they look at different directions to verify their identities.
“Facebook is working on Facial Recognition-based Identity Verification, asking users [to] take [a] selfie looking at different directions,” reads Wong’s tweet.
The new feature, which is currently in the testing phase, requires users to take a selfie as “we need to see your face at different angles to help us confirm you’re a real person.”
It instructs the users to hold their phones at the eye level, raising the phone so that users’ won’t be looking down at the camera as the AI is taking a photo of the user’s face. It then tells users to “follow on-screen” instructions by slowly turning their heads to the direction of the arrows for the camera to capture all the angles of the users’ face.
Users need to position their heads within a circle on their screen and move it according to the instruction, which is prompted upon completing one angle. Once all of the angles were captured, a confirmation message will be displayed.
“Video Selfie Complete. Thanks for completing this step. Submitting this video will help us confirm your identity,” reads the confirmation message.
As part of the disclaimer, the screen will also project a disclaimer stating that the video taken from the video selfie authentication feature will not be shared with any person and that it will automatically be deleted within 30 days after the selfie was taken.
In the same Twitter thread, Wong raised rhetorical questions relating to the feature and the propensity of the facial recognition technology to be used against someone’s privacy.
“If you were Facebook, how else would you make sure the user is real? How else would you ensure the integrity and prevent frauds which happen in e-commerce?” Wong asks. “What if FB just didn’t verify anyone’s identity and let fraud and misinformation campaigns happen? Would this be ideal?”
Facebook’s decision to test a facial recognition-based identity verification feature comes during a time when policymakers and privacy advocates have been slamming the use of facial recognition technology by law enforcement and private companies.
Many privacy watchdogs have called for the regulation of facial recognition and have slammed law enforcement agencies for using the technology with little to no regulatory buffer. They argued that the right of people to privacy supersedes the need for the use of the said technology.
In the past months, many states have already implemented regulations on the use of facial recognition by the police and private entities. It is spearheaded by San Francisco (where Facebook’s main office is located) after the city banned facial recognition use by the police and other government agencies.
It is still unclear whether or not Facebook will continue with its plan to roll out the facial recognition-based verification feature as of the moment.