Last week, the Metropolitan Police announced that it would be deploying facial recognition cameras during the Christmas shopping season in Soho, Piccadilly Circus and Leicester Square as part of a pilot program to test the technology. This announcement has understandably concerned many, and added fuel to the ongoing debate as to how facial recognition technology should be legislated and whether or not it should even be implemented. While some of these concerns are entirely justified, it is easy to blow them out of proportion and forget the very real benefits associated with the program.
Facial recognition surveillance has an understandably bad reputation. Depending on your background, its mere mention may conjure up intimidating images of either Person of Interest or a dystopian China, where an expansive surveillance system enables the government to encroach onto every aspect of their citizens’ daily lives. Indeed, there are two issues at the center of the facial recognition debate that need to be addressed for it to be implemented effectively: privacy concerns and its cost-effectiveness
For many, facial recognition is the ominous first step down a slippery slope that ends with the government having a full catalogue of biometric data on every citizen, and using it without any sort of legislative oversight. In the past, collection of biometric data was limited entirely to fingerprints and DNA samples, both of which were only accessed in the case of major crimes. Furthermore, both are heavily regulated, and in particular a warrant is required to compel a DNA sample.
Unfortunately, no such protections are in place for facial recognition technology, and many rightfully feel that its implementation, even as a pilot program, should be more heavily regulated so as to ensure that the privacy of individuals is not being violated. It is nonetheless true that facial recognition uses data that is available to anyone passing said individual on the street-so why are people concerned that their privacy is being violated? The government already knows what its citizens look like from their passports and driving licenses, what’s the difference with facial recognition technology?
The difference is as follows: facial recognition may not drastically change what information is being collected, but it does drastically change what the information can reliably be used for. With a fully functional facial recognition software, the government could theoretically track the whereabouts of all citizens wherever there is a CCTV camera. Considering that the UK has 20% of the world’s cameras for about 1% of its population, the implications are quite worrying. While this possible use of the technology in and of itself construes an unacceptable invasion of privacy for many people, stories of how China plans to use its surveillance network to enforce its social credit program demonstrate what a powerful tool for authoritarian control this technology can be if left unregulated.
Another concern brought up by those that oppose the development of this technology is that of how costly it is, relative to its actual effectiveness. After all, in May the technology was revealed to be only 2% accurate despite the fact that the Home Office put out an almost £5m contract for facial recognition software last year. Wasteful spending is bad enough, but when those false positives can then be used as the basis for falsely questioning and detaining individuals, there is real cause for concern.
Having established the privacy risks and relative ineffectiveness of facial recognition software, why tolerate its existence at all? The main consideration is that these issues all have to do with implementation, not necessarily with the technology itself, and of course entirely ignore the technology’s tangible benefits and helpful applications.
The privacy issues themselves are often overblown-while the hypothetical worst case scenarios are of course worrying, we are worlds away from descending aforementioned slippery slope. In the case of the pilot program for example, members of the public were allowed to opt out of being scanned and were presented with information on the technology and the program’s goals. It is crucial that facial recognition technology be properly regulated before it is developed further and potentially implemented on a larger scale, and educating the public is absolutely critical in reaching that goal. These programs, which are themselves heavily regulated, are therefore minimally invasive and serve the important task of bringing the issue into the public sphere.
While the ineffectiveness of the technology clearly shows that much work remains to be done in its development, those arguing that this means the government should stop investing in it have lost sight of the bigger picture. Whereas in the short term facial recognition will be very expensive to develop, the long run savings in terms of man hours and of course the increase in criminals apprehended means that the government will more than recoup its investment in the long term.
Unanswered questions nonetheless remain, notably about how to legislate this bleeding edge technology. But at the end of the day pilot programs such as the Met’s serve the important purposes of educating the public on its legislation and gathering technical data on how to continue to improve the technology so that it can best serve the country.