Meta will test facial recognition technology with more than 50,000 celebrities to prevent fraudulent ads - Photo: SHUTTERSTOCK
The Guardian reported on October 22 that Meta, the parent company of Instagram and Facebook, will test facial recognition technology with more than 50,000 celebrities worldwide from December, to prevent fraudulent ads from their images.
Accordingly, if Meta's management system detects an advertisement using celebrity images to scam, the advertisement will be deleted immediately.
“This technology will be faster and more accurate than manual moderation, allowing us to protect users from scams and protect celebrity images,” said David Agranovich, director of operations at Meta.
The company said initial testing with a small group has shown “promising results” in terms of speed and efficiency in detecting fraudulent ads.
During the initial rollout phase, celebrities will receive a notification on the app that they are selected to participate and can withdraw at any time.
Additionally, Meta also plans to use this technology to help users on Facebook and Instagram platforms recover hacked or locked accounts by uploading selfie videos.
Previously, in 2021, Meta stopped using facial recognition technology on its apps, especially in suggesting tagging users in photos, due to privacy concerns.
With the technology being reused after three years, Mr Agranovich stressed that the facial data will be deleted immediately after the test is completed and will not be used for any other purposes.
Facial recognition is reportedly just one of several measures Meta uses to detect fraud.
In recent years, Meta has faced pressure from politicians and regulators to tackle celebrity impersonation scams.
The company has been sued by mining magnate Andrew Forrest for failing to deal with scams using his image, and is also facing another lawsuit from the Australian Competition and Consumer Commission.
Source: https://tuoitre.vn/meta-thu-nghiem-cong-nghe-moi-chong-quang-cao-lua-dao-2024102215003112.htm
Comment (0)