Last October, Meta plunged his toe into the world of facial recognition – an area where he had a delicate record – with an international test of two new tools: one to stop scams according to the similarities of famous people and a second facial recognition function to help people take over the Facebook or Instagram compromise accounts. Now this test extends to another notable country.
After initially kept his facial recognition test in the United Kingdom, Meta Wednesday also started rolling the two tools there. And in other countries where the tools have already been launched, the protection of “celebrity baits” is extended to more people, the company said.
Meta said he had had the green light in the United Kingdom after “after committing himself to the regulators” in the country-who himself made a big point in the last months of AI adoption. No word on Europe yet, the other key region where Meta has not yet launched facial recognition tools.
“In the coming weeks, public figures in the United Kingdom will begin to see integrated notifications to let them know that they can now choose to receive the protection of celebrity baits with facial recognition technology,” said a statement from the company. The optional tools of this “verification of video selfies”, said Meta.
When Meta described for the first time the deployment of the two new facial recognition tools, the company said that features were not used for something other than the objectives described: fighting advertising and user verification.
“We immediately delete all the facial data generated from ads for this unique comparison, that our system finds correspondence, and we do not use it for other purposes,” wrote Monika Bickert, VP of Meta content policy in a blog (which is now updated with the details of British expansion).
However, developments are involved at a time when Meta goes all on AI in her business.
In addition to building his own language models and using AI on his products, he would also work on an autonomous AI application. He also intensified lobby Around technology, and taking into account his two hundred, what he deems is risky AI applications such as those that can be armed (the involvement being that what Meta built is not risky).
The company’s decision to create tools that solve immediate problems on its applications is probably the best approach to acquire any new facial recognition feature. These features correspond to this invoice. As we have said before, Meta has been accused for many years of not having prevented the crooks from diverting the faces of famous people in order to use his advertising platform to run scams such as doubtful cryptography investments to unrestricted users.
Facial recognition has been one of Meta’s more thorny areas over the years when it has worked with AI technology. More recently, the company last year agreed to pay $ 1.4 billion To settle a long -standing trial in Texas, where it was continued on a collection of inappropriate biometric data linked to its facial recognition technology.
In 2021, Facebook closed its old -fashioned facial recognition tool for photos, a functionality that had been faced with a number of regulatory and legal problems. Interesting fact, at the time retained Part of the technology: its Deepface model, which he declared to integrate into future technology. This could well be part of what is built with today’s products.
