The face-recognition app that will push tech companies into a privacy storm
02 May 2016
In this fantastic technologically-enabled world, we are treading so often so close to the line on privacy, we often stumble over it.
Dating app Tinder was the latest to find this out, as it got into trouble trialling its new friend-finding feature amongst some users in Australia. The feature automatically displayed to people which of their Facebook friends are also using Tinder. It’s supposed to help them to group together for more social things.
Turns out not everyone wants their Facebook friends to know they’re on Tinder. And it’s one of those apps where the privacy guarantee is rather important. At the very least, people have been arguing, the feature should be opt-in, not opt-out. But that, of course, would mean very few people would actually get any value using it. And so the debate rages on.
Meanwhile, whilst mainstream companies like Tinder, and Facebook and Google, tippy-toe around around these difficult issues, trying to push what’s possible within a delicate curve of what’s desirable, one app maker in Russia has seen the line and marched straight over it.
FaceFirst produces an app that, in Russia, can be given a photo of someone taken in the street, and can match them with high accuracy to their profile on VK.com, the Russian equivalent of Facebook. Only last year, privacy experts had declared such capability to be far away. Not so far away after all. With suitably high quality pictures, the service records a 9 out of 10 success rate. With pictures taken in the street, that might drop to 7 out of 10.
It has some great beneficial applications, of course. Robbers caught on CCTV? Identified in a flash.
But it has to be doubted that the hundreds of thousands of people that have downloaded the app to date have such altruistic ends in mind.
The power of the app to cause harm was soon demonstrated, when some people decided to use it to ‘out’ a number of Russian porn stars.
FindFace’s founder Maxim Perlin said that the company was making every effort to protect users from malicious acts, but admitted there was little he could do to stop the service being used this way.
Let’s just map how this is likely to go.
The app is launched in America and Europe (although it might get blocked more energetically in Europe. At the moment, Facebook keeps its own face recognition programme from European users because of EU privacy concerns).
It is an overnight success. In a very short time, some creatively horrible use cases hit the media. We could speculate, but they might include:
- Paedophiles getting photos of parents dropping their kids off, getting their details, so that they can persuade kids that they have been sent by their parents to pick them up later
- Random women being stalked by people who take their photo in the street and get their details, and want to show off how much they know about them
- Police or trading standards officers, who could be traced and targeted by criminals under threat of prosecution
I’m sure that whatever it would be that becomes the catalyst, it would be something else, as there are so many possibilities. It would probably be something to do with celebrities, although I can’t quite think of the angle.
In any case, something happens that makes the front pages. So then the tech companies come under fire. After all, even if the app makers are based in a jurisdiction that makes them hard to reach, they rely on the platform providers. So Apple and others will ban the app from their store. That might be that for the Apple universe, but not so for those less tightly locked down.
So then the tech companies will be challenged to be more proactive in ensuring that such software cannot run on their devices. The problem, of course, will be that although you could make it difficult for regular users to get hold of it, for those seriously intent on mischief, there will always be a way.
You know, serious criminals and abusers. Oh, and the FBI, of course. According to a recent freedom of information request, the FBI’s standard warrant to social media companies like Facebook seeks copies of all images you upload along with all the images you’re tagged in. This sort of technology will be the obvious additional element to such requests.
The tech companies are caught in a genuine dilemma on this one. They want to be able to introduce new services that make amazing things possible for regular users who are trying to track down old friends, or who want it easy to tag all their friends in their pictures so they know they can find them again later.
But they also know the potential for abuse. For instance, Google – which got plenty of kick-back as it was over the potential for wearers of Google Glass to be videoing people without their knowledge – did see the dangers enough to ban facial recognition apps on that device.
But they are reluctant to be pushed into being too timid, to the frustration of privacy campaigners, who have walked out of talks before now frustrated by what they saw as industry obstinancy.
This isn’t a trivial matter for them. They know they could potentially earn a lot of money from advertisers who can push individualised messaging to people who are recognised on their way into a store. We’ve seen the scenario in movies. It’s just a heartbeat away.
But then that’s when your values come to the fore, when committing to a principle like user privacy actually involves giving up significant revenue opportunities.
Over the next five years, get ready to watch those values being tested. I predict that Apple will quickly take action. They don’t get significant revenue from advertising, and they distinguish themselves from the others by not selling their customers’ data as part of their business model.
The others are really going to struggle.
And within the next decade, we will all understand a little bit better that we are all living our lives in public from now on.
