Meta announced the closure of Facebook’s facial-recognition system in the coming weeks. The social network will also delete individual biometric templates for more than a billion people.
When publishing new images, the platform will stop suggesting tagging of people in those images. It will also stop sending notifications to users when someone posts a photo or video featuring them.
The decision was taken amid growing concerns over broader use of facial recognition technology, including a lack of regulatory oversight.
“In the face of ongoing uncertainty, we consider it prudent to limit the use of facial recognition to a narrow set of applications,” said Jerome Pesenti, the company’s vice president of AI.
He said that facial recognition can still be useful in certain situations, such as regaining access to a locked account. However, the company concluded that a privacy-focused approach is more appropriate than a “broad outward-facing strategy”.
“Every new technology brings both benefits and public concern. We will strive to find the right balance,” Pesenti said.
The image-description system for the visually impaired will also be affected: it will no longer name people detected in photographs.
Facebook’s facial-recognition system first appeared in 2011. It used to analyze images and identify people in them without their consent. According to Pesenti, numerous user complaints about this also influenced the company’s decision to limit the use of the technology.
Earlier, Facebook carried out a major rebranding, renaming the company to Meta. The social network, however, kept its former name. The branding occurred amid a scandal surrounding the leak of internal company documents.
In October, former Facebook employee Frances Haugen testified before the U.S. Senate that the social network uses algorithms to inflame hatred for profit.
A week later, Nick Clegg, the company’s vice president for global affairs, said that the corporation is prepared to grant regulators access to AI models to demonstrate their harmlessness.
In mid-October, it emerged that Facebook’s algorithms remove only a portion of content that violates the platform’s rules.
Follow ForkLog’s news on Telegram: ForkLog AI — all the news from the world of AI!