Facebook says it’s giving the ax to one face-scanning tool and introducing a wider rollout for another — this time by giving users a heads up about it first.
The tag suggestions setting — or the one that allows users to easily tag their friends in photos based on Facebook-generated name recommendations — will no longer be available, according to a new blog post by Srinivas Narayanan, an applied research lead of Facebook AI.
Instead, everyone will now have the face recognition setting that the company initially rolled out to some users in 2017, and which the company has claimed is used to beef up user security, such as by alerting users when they may appear in a photo they’re not tagged in or when someone may be using a user’s photo on their own profile.
Narayanan noted that with Facebook’s rollout of the face recognition tool to those who did not previously have it, they’ll see a notice in their News Feed about how the tool works and how it’s used by the company. The notice will also include a button to turn the feature on or off, and Facebook says it will not enable face recognition by default for folks who are just getting it and don’t elect to participate.
“If you do not currently have the face recognition setting and do nothing, we will not use face recognition to recognise you or suggest tags,” Narayanan said. “In addition, features like Photo Review, which lets you know when you appear in photos even if you are not tagged, as long as you have permission to see the post based on its privacy setting, will not be activated. People will still be able to manually tag friends, but we won’t suggest you to be tagged if you do not have face recognition turned on. If you already have the face recognition setting, you won’t receive a notice.”
Tag suggestions was used to suggest names for users’ friends on Facebook. Facebook said that users who had the feature enabled will now have facial recognition by default and will be notified about the switch and how it works.
If all of this transparency about data privacy and education about its policies sound uncharacteristically altruistic, think again. Part of the reason that Facebook decided to roll this update out now might have something to do with the massive $US5 ($7) billion settlement it reached with the U.S. Federal Trade Commission in July (though it is still pending approval).
When asked for comment on the timing of the announcement, a spokesperson told Gizmodo:
This work is the last phase of a long-planned deprecation of the Tag Suggestions setting. For context, the 2017 effort was the culmination of over two years of work. During this time we’ve worked to improve the rollout experience based on feedback from people, privacy stakeholders, and regulators.
The FTC order is still pending approval but per our earlier post, we’ve agreed to a more comprehensive privacy framework that governs how we build our products. Face recognition is part of that conversation.
As part of an FTC complaint about the company, the agency alleged that Facebook violated user privacy in myriad ways, including by misleading users about its facial recognition policies practices and — surprise! — the company’s tag suggestions feature, which was turned on by default on user accounts.
Ultimately, Facebook was ordered to adhere to a number of new measures around its face recognition technology, including educating users on how it works and requiring the company to obtain their consent if their data is used for reasons beyond what was explicitly outlined by the company.
“Long-planned” or not, notifying users about the use of a technology that many of them may not have understood years after the fact seems about right for a company that continuously botches its handling of user data.