Another Day, Another Company Scraping Photos To Train Facial Recognition AI


If your face can be found online, chances are it’s now part of a facial recognition database. These aren’t the ones being utilized by law enforcement, although those are bad enough. The ones used by law enforcement are littered with millions of noncriminals, all part of a system that works worse than advertised 100% of the time.

The faces aren’t in those databases (yet!), but they’re being used to train facial recognition AI with an eye on selling it to law enforcement and other government agencies. Another photo storage company has been caught using users’ photos to fine tune facial recognition software… all without obtaining consent from those whose faces became fodder for the tech mill.

“Make memories”: That’s the slogan on the website for the photo storage app Ever, accompanied by a cursive logo and an example album titled “Weekend with Grandpa.”

Everything about Ever’s branding is warm and fuzzy, about sharing your “best moments” while freeing up space on your phone.

What isn’t obvious on Ever’s website or app — except for a brief reference that was added to the privacy policy after NBC News reached out to the company in April — is that the photos people share are used to train the company’s facial recognition system, and that Ever then offers to sell that technology to private companies, law enforcement and the military.

This has been 2019’s theme for the first five months of the year. Users of popular photo apps and services are being notified belatedly — and not by the companies performing the harvesting — that their faces are an integral part of law enforcement machinery and/or the military-industrial complex.

Ever’s oh-shit-we-got-caught statement doesn’t offer much mollification.

Doug Aley, Ever’s CEO, told NBC News that Ever AI does not share the photos or any identifying information about users with its facial recognition customers.

Lots of people would rather not be participants in creating surveillance tech. Most never seek employment at companies crafting products for law enforcement, intelligence agencies, and the US military. Without being informed, the photos they thought they were harmlessly sharing with family and friends have been used to make surveillance easier and more pervasive, if not actually any better.

Ever is just the latest. Prior to this, Flickr photos were swept up in a facial recognition data set compiled by IBM.

The photo is undeniably cute: a mom and a dad — he with a stubbly beard and rimless glasses, she with choppy brown hair and a wide grin — goofing around and eating ice cream with their two toddler daughters.

The picture, which was uploaded to photo-sharing site Flickr in 2013, isn’t just adorable; with a bunch of different faces in various positions, it’s also useful for training facial-recognition systems, which use artificial intelligence to identify people in photos and videos. It was among a million images that IBM harnessed for a new project that aims to help researchers study fairness and accuracy in facial recognition, called Diversity in Faces.

IBM also apologized for using people’s photos for its data set without their permission. It said users were welcome to opt out at any time, but did not give users tools to find out whether their photos had been used. Nor is there any way to expeditiously remove found photos other than by handing over your Flickr ID to IBM.

And if it’s not a tech company harvesting photos to run AI tests, it’s random internet users showing just how easy it is to compile a data set using other people’s photos.

Tinder users have many motives for uploading their likeness to the dating app. But contributing a facial biometric to a downloadable data set for training convolutional neural networks probably wasn’t top of their list when they signed up to swipe.

A user of Kaggle, a platform for machine learning and data science competitions which was recently acquired by Google, has uploaded a facial data set he says was created by exploiting Tinder’s API to scrape 40,000 profile photos from Bay Area users of the dating app — 20,000 apiece from profiles of each gender.

The data set, called People of Tinder, consists of six downloadable zip files, with four containing around 10,000 profile photos each and two files with sample sets of around 500 images per gender.

Tinder’s reaction was to call this a violation of its Terms of Service. But this determination doesn’t undo the damage nor make it impossible for someone else to do the same thing. Tinder users spoken to by TechCrunch weren’t happy their photos — some of which have never been uploaded outside of the app — are being used by a person they don’t know to perform research.

It’s not that there’s no legitimate uses for publicly-available photos. But transparency is the key and no one harvesting photos to train AI systems or perform research seems too concerned about being upfront with the people whose photos they’re using. It’s even worse in the case of Ever, where the app company itself is the one developing facial recognition software on the side, which should make users question the intent of the app developers. Did they really want to offer another photo service or were they just using this to gather faces for their real moneymaker?

Filed Under: , , ,
Companies: ever

Rate this comment as insightful
Rate this comment as funny
You have rated this comment as insightful
You have rated this comment as funny
Flag this comment as abusive/trolling/spam
You have flagged this comment
The first word has already been claimed
The last word has already been claimed
Insightful Lightbulb icon Funny Laughing icon Abusive/trolling/spam Flag icon Insightful badge Lightbulb icon Funny badge Laughing icon Comments icon

Comments on “Another Day, Another Company Scraping Photos To Train Facial Recognition AI”

Subscribe: RSS Leave a comment
Crinisen says:

IBM also apologized for using people’s photos for its data set without their permission. It said
users were welcome to opt out at any time, but did not give users tools to find out whether their
photos had been used.

Obviously they need a way to figure out if you are the person in the photo, so please turn on your web cam and slowly turn your head left, right, up, and down as far as you can…

Add Your Comment

Your email address will not be published. Required fields are marked *

Have a Techdirt Account? Sign in now. Want one? Register here

Comment Options:

Make this the or (get credits or sign in to see balance) what's this?

What's this?

Techdirt community members with Techdirt Credits can spotlight a comment as either the "First Word" or "Last Word" on a particular comment thread. Credits can be purchased at the Techdirt Insider Shop »

Follow Techdirt

Techdirt Daily Newsletter

Techdirt Deals
Techdirt Insider Discord
The latest chatter on the Techdirt Insider Discord channel...