The rapid rise of facial recognition tech has sparked an arms race of countermeasures in research labs and at street level – here’s why.
You know the debate about facial recognition has become serious when The Economist wades in.
Last week the magazine ran an article looking at how face recognition has become more prevalent, from smartphone security systems and social media to supermarkets, billboards and law enforcement agencies. The latter has arguably generated the most discussion of late, or at least is where much of the subsequent backlash has been directed.
This is certainly the case here in Hong Kong, which has generated a lot of media attention over the spectacle of anti-extradition law protesters using laser pens during street standoffs with police, purportedly to thwart facial recognition software (although you’d think wearing helmets and gas masks would be just as effective).
The facial-recognition backlash and subsequent arms race has been going on for several years now, but as more organizations adopt the technology, the more the arms race arguably matters. Face recognition is fast becoming one of the most disruptive technologies out there, in both good and bad ways, and it’s being rolled out faster than society in general is prepared to deal with.
That’s partly because the technology is often invisible to most of us when we encounter it. Using FaceID to unlock your iPhone is an obvious use case – a retailer using it in a store to detect if a customer is unhappy or pitch personalized ads at a soft drink dispenser is not so obvious. And of course facial recognition used for security and surveillance apps is intentionally behind the scenes. Giscle has a whole list of potential use cases across a multitude of sectors here – and most of them involve apps where the facial-recognition element would go unnoticed by the person whose face is being scanned.
This is why lots of people find facial recognition more frightening than other biometric technologies. Using a fingerprint or retina to unlock a device or gain access to a facility requires you to actively interface with the scanning device – you know it’s there, why it’s there, and what the result will be if it works properly.
Facial recognition is a different story, depending on the app. We know it exists, but most of us only have a vague idea of what it’s being used for, apart from spotting terrorists and other criminals – which means that many people associate the technology only with those uses, which makes other relatively benevolent use cases sound scary or Big-Brotherish. It doesn’t help that facial recognition has a spotty track record out in the field, to include issues with racial bias.
Little wonder, then, that researchers are looking at how facial recognition can be thwarted. There are various techniques, but my favorite (apart from being a fan of Insane Clown Posse) is a baseball cap equipped with tiny LED lights that project infra-red dots onto your face. The Economist writes:
Many of the cameras used in face-recognition systems are sensitive to parts of the infra-red spectrum. Since human eyes are not, infra-red light is ideal for covert trickery.
(Ironically, two of the backers of that project are Chinese University of Hong Kong and China-based Alibaba, which amuses me no end, all things considered.)
While the face-recognition arms race is fascinating, the real story here is the fact that it exists at all. The backlash speaks volumes about how ready we are for mainstream use of this technology. Which is to say, not very.
It’s also important to remember that facial recognition tech is a subset of video surveillance technology using analytics and other tools to determine what people are doing in videos – exiting through an entrance door, lying down in a public space or shopping in a cashierless grocery store, for example. Combine all of this with the data already being collected and compiled into digital profiles of what we do both online and offline, along with cheap storage and more efficient compute power, and video surveillance is an extremely powerful and disruptive technology even without the ability to identify faces.
According to security expert Bruce Schneier, that’s why some city governments in the US have either passed or are considering laws to ban the use of facial recognition tech – at least temporarily while we pause and consider what kinds of policies we want to put in place that can govern the use of such technologies without unnecessarily stifling innovation.
But that’s not going to stop R&D, and it’s probably too much to hope for the same level of caution by at least some governments in this part of the world (China being literally the biggest example of a government gleefully embracing and weaponizing facial recognition as a social control mechanism).
Hence the backlash and subsequent arms race. But then most of our recent technological innovations have happened either in parallel or way ahead of any meaningful social discussion on the pros and cons. Like it or not, we’re likely to have facial recognition thrust upon us and figure out how to live with it afterwards.
Or invest in high-tech baseball caps. Here’s hoping they’re easily upgradeable.