> The change affects more than a third of Facebook’s daily users who had facial recognition turned on for their accounts, according to the company.<p>I'm surprise the number is that low. Wasn't enabling this a default?<p>> That meant they received alerts when new photos or videos of them were uploaded to the social network. The feature had also been used to flag accounts that might be impersonating someone else...<p>That's actually a pretty valuable and legitimate use case for such a system, even from a very pro-privacy perspective.<p>> Although Facebook plans to delete more than one billion facial recognition templates, which are digital scans of facial features, by December, it will not eliminate the software that powers the system, which is an advanced algorithm called DeepFace. The company has also not ruled out incorporating facial recognition technology into future products, Mr. Grosse said.<p>It's also almost certainly not going to delete the face-specific image tagging either, which means they could re-derive everything they're "deleting" relatively easily.
The facial recognition feature is how I figured out someone was doxxing and harassing me on their page. Getting the notification that it thinks I appeared in a photo they had ripped from my FB profile was the tip off. I have mixed feelings.
So sad that we can't have nice things. I like the feature that automatically tags people in pictures.<p>Sometimes I have trouble remembering people's names (not face blindness or similar, just regular trouble I guess). I would love to have smart glasses that show the name of someone I've met before but don't quite remember. But alas, our society being what it is people will abuse facial recognition, so I can't have that it seems...
It seems this is likely a result of the 2019 FTC Consent Order on Facebook [1]. Facebook has had to implement privacy processes which are monitored by a 3rd party who reports to a judge and facial recognition was one of the issues sparking the order.<p>The consent agreement [2] has whole sections and rules specifically for facial recognition data.<p>Likely a review program and criteria (approved by the monitor) were set up based on these rules, and when the tagging feature was pushed through this review it failed. Rather than risk the judge sanctioning them for failing to meet the order, they decided shut down the feature.<p>[1] <a href="https://www.ftc.gov/news-events/press-releases/2019/07/ftc-imposes-5-billion-penalty-sweeping-new-privacy-restrictions" rel="nofollow">https://www.ftc.gov/news-events/press-releases/2019/07/ftc-i...</a><p>[2] <a href="https://www.ftc.gov/system/files/documents/cases/c4365facebookmodifyingorder.pdf" rel="nofollow">https://www.ftc.gov/system/files/documents/cases/c4365facebo...</a>
Their announcement is at <a href="https://about.fb.com/news/2021/11/update-on-use-of-face-recognition/" rel="nofollow">https://about.fb.com/news/2021/11/update-on-use-of-face-reco...</a>.<p>(via <a href="https://news.ycombinator.com/item?id=29084199" rel="nofollow">https://news.ycombinator.com/item?id=29084199</a>, but we merged that thread hither)
> We need to weigh the positive use cases for facial recognition against growing societal concerns, especially as regulators have yet to provide clear rules.<p>Disclaimer: I have never owned an Apple product, and I haven't used Facebook for roughly a decade.<p>Apple have been doing face recognition for a while now, arguably at a similar scale, but their use case is pretty straightforward. With FB face recognition is not the real concern, it's the fact they have been caught misusing personal data multiple times.
Here's where Facebook's brand reputation sits in my mind today:<p>I do not believe them at all. I do not believe they've had some change of heart regarding "many concerns about the place of facial recognition technology in society." I do not believe they intend to not use this technology in some other privacy-invading way.<p>Trying to figure out their real angle here. Avoid regulation? "Good" PR at a time when they desperately need it? A broader existential threat that I can't even fathom at the level of power their company currently wields in society?<p>Probably not much at this point they could do to change my absolute negative opinion of the company though, so there's that.
I get all the privacy concerns, but this was actually super useful for people who opted-in. Manually retyping everybody's name when tagging them in 50 photos from an event is torturous.
The hoops they have to jump through just to enable some really cool features down the line is sad.<p>I don't know why adding another opt in wasn't enough, it's not like the haters are going to believe they deleted the templates anyways.
Pity. I liked the feature. Fortunately, we’ve mostly all moved over to use Google’s shared albums with automatic facial recognition and sharing. My friends all have photos I take of them and likewise they share theirs with me.<p>I think Google has figured out a lot of fun sharing tools for people in trusted environments: location sharing, photo tagging, etc.<p>I’m rather impressed with it all. Good stuff.
A legitimate question I’ve always had when we read such announcements is how can we be certain they’re actually doing whatever they’re claiming / planning to do. Unlike software /OS modifications, it’s not feasible to somehow reverse engineer FB’s backend and find out if anything unusual is happening. At least without any attempt of compromising it.<p>Considering how powerful these orgs are, with low probability of facing any legal repercussions about potentially false claims, do we have any other option than to take their word for it?
Meta has no plans to discontinue the use of facial recognition though: <a href="https://www.vox.com/recode/22761598/facebook-facial-recognition-meta" rel="nofollow">https://www.vox.com/recode/22761598/facebook-facial-recognit...</a><p>The story of how Facebook's facial recognition capabilities came to be is an interesting one: <a href="https://en.wikipedia.org/wiki/DeepFace" rel="nofollow">https://en.wikipedia.org/wiki/DeepFace</a>
<a href="https://en.wikipedia.org/wiki/Face.com" rel="nofollow">https://en.wikipedia.org/wiki/Face.com</a><p>I had no idea that Face.com's algorithm was/is able to ID ~97% of all faces 'in the wild'. Nor did I know that Face.com had opened its API to use by dozens (hundreds?) of companies back in 2007. I wouldn't doubt if there are knock-offs / clones / reverse-engineered versions of the original Face.com API floating around. API security and design - in 200x - wasn't what it is today.
Highly unlikely that they will delete it. The info collected is too valuable for them to delete it. They may retain the hash or other unique identifiers even if they delete originals. They have shown again and again that they can't be trusted.<p>trust, once lost, can never be regained.
Getting rid of its facial recognition system is a good step, but lest people either forget or miss...<p><pre><code> Although Facebook plans to delete more than one billion facial recognition templates, which are digital scans of facial features, by December,</code></pre>
<i>it will not eliminate the software that powers the system,</i><p><pre><code> which is an advanced algorithm called DeepFace.
</code></pre>
<i>The company has also not ruled out incorporating facial recognition technology into future products,</i><p><pre><code> Mr. Grosse said.</code></pre>
Sorry for me not believing this. They’ve been caught lying so many times trying to save face that I fully expect them to double down on facial recognition after this announcement.
If it had been opt-in from the getgo, I think it would have been less controversial.<p>I still remember a time when it was common courtesy to <i>ask</i> someone before you took a picture of them.
Facebook facial recognition never actually learnt how to differentiate my face from my relative's face. It was cool since I had no "tracking" for them to save.
I am for this decision, but I don't understand their justification for it. What was the actual reason they pulled back on this? The US government is assuredly for maximum surveillance, so long as the publicly visible part is done by private companies that can be subpoenaed for their data. FB needs to stay in the good graces of the military-industrial-political complex to avoid anti-trust action.<p>FB doesn't do anything out of the good of their heart, so what are they getting out of this? I really doubt that winning the hearts of civil libertarians is at all important to them.<p>(I also noticed how the NYT managed to get a dig into the discredited China story when discussing a company that was surveilling billions of people worldwide using facial recognition technology.)
It's always confusing to see what the internet police go after.<p>Total scams ripping thousands to hundreds of thousands off from people (tech support scams etc). Crickets.<p>An opt in feature on the facebook platform that is useful (at least to some). Multiple Attorney General press conferences :)
Slightly off-topic but I deleted my Instagram yesterday. I was really holding out for one person who almost refuses to talk to me on any non-FB property but...oh well. I'm glad to be rid of it, but I can't shake that it's somewhat meaningless and I'll miss the many content creators I followed who have earned my business. I wonder if there's some way to follow them in a way not tied to my identity.
"We are never ever ever getting back together." Facebook user base quoting 21 century poet laureate.<p>Hey you remember all those times we analyzed you photos, and those times we did things we said we wouldn't? We totally aren't going to do those now. We are "Meta" not the evil Facebook, trust us. "Meta" has never lied to you, right?
Luddites. You can't un-invent technology. Like all technology, facial recognition has tremendous potential for good. We should learn to live with the technology, not attempt to forbid its use.<p>Also, Meta is a bunch of cowards nowadays. They should not have kowtowed to angry Twitter activists. It sets a bad precedent.<p>I don't like this system where activists bully big companies into compliance with their idiosyncratic moral systems instead of using the government and the laws to codify restrictions and rules in a transparent way that everyone can understand and in which everyone has a say.<p>If facial recognition is so bad, it should be illegal. If it's not illegal, FB has a responsibility to its shareholders to maximize profit within the law. If, as an activist, can't get support for making facial recognition illegal, then maybe it's not quite as bad as you claim.