Another convincing deepfake app goes viral prompting immediate privacy backlash

Image: @AllanXia

Zao, a free deepfake face-swapping app that’s able to place your likeness into scenes from hundreds of movies and TV shows after uploading just a single photograph, has gone viral in China. Bloomberg reports that the app was released on Friday, and quickly reached the top of the free charts on the Chinese iOS App Store. And like the FaceApp aging app before it, the creators of Zao are now facing a backlash over a perceived threat to user privacy.

Twitter user Allan Xia posted a neat demonstration of what the app is capable of yesterday with a 30 second clip of their face replacing Leonardo Dicaprio in famous moments from several of his films. According to Xia, the clips were generated in under eight seconds from just a single photograph, however Bloomberg notes that the app can also guide you through the process of taking a series of photographs — where it will ask you to open and close your mouth and eyes — to generate more realistic results.

According to Xia, the app only offers a limited number of clips for you to insert your face into. The app’s developer has likely trained their algorithms on each of these clips to easily re-map a user’s face onto them, as Xia speculates. The app cannot map your face onto any video clip of your choosing.

The technology looks similar to what we’ve seen recently from researchers at London’s Imperial College, who showed off technology that’s able to turn a single photo into a singing portrait. The difference here is that Zao is inserting your likeness into an existing video, rather than animating a photo of you directly. Nevertheless, it demonstrates how quickly the underlying technology has evolved: what once required hundreds of images to create a rather convincing deepfake video now requires just a single image with better results.

Zao’s privacy policy generated an almost-immediate backlash from users, who bombarded its App Store listing with thousands of negative reviews. The Zao app lists the developer as Changsha Shenduronghe Network Technology, which Bloomberg notes is a wholly owned subsidiary of Momo, a Chinese company that owns a live-streaming and dating service.

The privacy policy includes a clause which says that its developer gets a “free, irrevocable, permanent, transferable, and relicense-able” license to all user-generated content, according to Bloomberg. The company has been forced to quickly respond to the criticism, and now says it won’t use its users’ photos or videos for anything other than app improvements without their consent. It will also erase user data from its servers when users delete their data from the app.

It’s a similar controversy to the one that surrounded FaceApp earlier this year, when the face-aging app again went viral in July. The app’s developer was forced to clarify its privacy policy, and to offer users the option of deleting their photos off its servers if they wished. In the case of FaceApp, commentators were quick to point out that the app’s privacy policy was no more invasive than many of the most popular mobile apps across the world.

Protesters in Hong Kong are going to great lengths to cover their faces over fears that police are using facial recognition technology to identify and arrest targets. People are increasingly aware of how important their facial imagery data is, and are rightly concerned about companies who don’t make adequate safeguards to protect it.

Whenever a service is provided for free, a company is inevitably profiting from your data. Sometimes it’s for better ad targeting, sometimes it’s to train their AI for better facial recognition. You often don’t know.

Comments

I think we have little choice but to brace for this to take over our lives, like Facebook did, or Google did and see how we cope. Trying to ban such poop or strongly regulate it will only make it more coveted!

I disagree and think deep fakes are very dangerous. Not sure if your in the United States or not but there is an election coming up soon and we have a president who would use something like this to his advantage. So much so that the opposing party has warned deep fakes can create a minefield.

Yep, the righteous opposition on the left wouldn’t use this to their advantage at all. Nope, not the perfect progressives who only want what’s best for this country…

At least the left doesn’t get angry at Puerto Rico during hurricane season via twitter and treats the island as if they keep asking for hurricanes. Anyway tech like this is a weapon of danger that can be used in politics, journalism, etc.

I’ll leave this here also when someone is president HE or SHE doesn’t get to pick the color of the folks that they represent. Keep that in mind.

Puerto Rico had emergency relief almost immediately while New Orleans was left to drown. Comparatively, who would you like in charge of declaring an emergency?

I would like someone who doesn’t get an attitude at the mere mention of Puerto Rico having an hurricane in charge. Here is the exact quote "Wow! Yet another big storm heading to Puerto Rico. Will it ever end?" from his tweet.

And yet I don’t see anywhere in the tweet where he is blaming Puerto Rico or its citizens at all. I think you are inferring things that simply aren’t there, which says more about you than anything else.

Puerto Rico is like 75% white. Not a fan of Trump at all, but I don’t get why people think the Puerto Rico stuff is racial.

Really? You need an explanation? Here it is: Language.

Luckily those on the right are too hamfisted to make anything even close to plausibly believable even with this technology.

Deep fake news!

Why don’t they use cryptographic watermarks in political videos then? Same as they do for security footage. Editing the footage would break the watermark and is easy to prove the video is fake then.

These gimmicky apps are fun for about 3 seconds. I don’t understand the appeal of these apps to begin with.

Do you / did you also like snapchat dog filters?

If not, then you probably aren’t the target market, and that’s okay

Correct, I am not a target market for this, but with everything going on today with our privacy, people Should really be asking themselves if using Something is worth the risk given the Minimal amount of entertainment it returns. How often would one use this app? Once, twice? Maybe a handful of times? Is that worth giving your information to some company?

Again, there is a target market they are going for that don’t think are about these things. I agree it would be great if people did think about it more but it would also be great if people didn’t get themselves into credit card/gambling debt, the world is filled with traps aimed at specific types of people, not everyone who gives their photo away or gambles is going to be impacted by it.
Take smoking as well, since the 80s, no one has believed that smoking was healthy. People now know it kills you and there are still millions of people that do it for enjoyment.

Tl:Dr
Short term fun vs risk of long term loss is a hard equation

We will be hearing a story soon about how this app steals info from your phone.

I, for one, do not care if someone has a picture of my face. What the heck could they do with it… better ad targeting? Good, the ads I get currently SUCK. Please give me better ones

Who said anything about ad targeting? Think a little harder. Blackmail is a giant concern with this sort of thing. Think about your run of the mill scams where someone in Nigeria emails your relatives saying you’re stuck in a prison somewhere and they need to send 10 grand to them to release you. Now imagine how much more powerful that scam becomes when your grandparent receives a video of YOU actually in jail somewhere pleading for your release, when in reality you’re at work or out with friends. Other obvious ways for this to ruin people’s lives are through deepfake porn videos of you sent to coworkers, and political impostors. The implications of this are staggering. We haven’t even began to scratch the surface.

You hit the nail on the head. You don’t even have to be too much of a conspiracy theorist to know how this could set people up…we have video evidence of you committing the crime, but that wasn’t me! How will we be able to separate fact from fiction?

Even more scary when you add the faking of audio (like the Joe Rogan example). Eventually all they’ll have to do is get you to say "hello" when answering a call to get the needed sample.

All the more reason to better educate people about how to safely use the Internet then really.

If you’re not doing the basics to check something like that by say… calling them first to check if they’re in prison then the saying about fools and their money springs to mind.

Perhaps an Internet license is required? You installed the Yahoo toolbar? You gave your bank details to that nice Nigerian prince? Sorry that’s a fail, you’ll have to retake the test.

Except all of this has already been possible for a really long time, you just had to do it manually before.

When Photoshop came around for editing pictures people complained about his same thing and here we are today and it hasn’t really been a problem.

The hard part is so much of our older generations are barely or still catching up with the world as it is now and something like this will throw so much of the populous into absolute chaos not being able to understand. Lives will be ruined before people can catch up to the modern version of the dopey scams of yore.

I thought social platforms had already turned people into deepfakes, do we really need an app over and above that?

View All Comments
Back to top ↑