Sir David Davis writes for The Telegraph on the potential risks posed by live facial recognition cameras


As published on The Telegraph:

Everyday millions of people walk past millions of cameras capturing billions of images. Unless you live in the most remote parts of the Scottish Highlands it is near enough impossible to avoid being caught on CCTV.

The UK has had these kinds of systems for decades now, and they have become a part of everyday life. Go to the shops, out for a coffee or down to the pub and you will inevitably be picked up by multiple cameras. On average a Londoner is captured by CCTV some 300 times per day.

With an estimated 6 million surveillance cameras in the UK, we are the most surveilled nation in the democratic world – beaten only by China. With one per cent of the world’s population we have 20 per cent of the world’s CCTV cameras.

It is broadly accepted that surveillance cameras are a necessary tool for the police to fight crime and for businesses and homeowners to protect their property. However, it must be debated whether we are right to put quite such reliance on such an intrusive technology.

Look at the proposed use of live facial recognition cameras. This is a technology that has sprung up with no legal backing and without consulting the public.

The Met and other forces have been trialling the technology for years now. These live facial recognition cameras allow police forces to observe large groups of people with footage being automatically checked against databases containing a ‘watch list’ of photographs of wanted offenders. In theory, the system alerts officers when a suspect is matched to a photograph.

Such technology has already monitored music concerts, Christmas shoppers and football fans attending the 2017 UEFA Champions League Final in Cardiff. The police justify this as an effective tool in identifying wanted suspects who pose serious risks to the public.

However, as we have seen from the trials, the system is far from reliable. The Metropolitan Police have deployed the cameras on eight occasions. When they did it resulted in an error rate of 93.33 per cent – in other words, of the 150 people picked out by the cameras, 140 were innocent people. On 30 occasions where South Wales Police used the technology, the cameras picked out 2,784 people. 2,507 of those identified by the cameras were innocent bystanders – representing an error rate of 90 per cent.

These are alarming error rates. If the police are allowed to roll out these cameras as a routine surveillance tool it will lead to hundreds of innocent people being stopped by the police.

This is not a hypothetical scenario. It has already happened. A 14-year-old child was stopped and fingerprinted by police after misidentification by facial recognition cameras. Innocent members of the public have been stopped and ID’ simply because the cameras could not identify them when wearing hooded coats and scarves. One man, who objected to having his face scanned, was fined for a public order offence after avoiding the cameras. These are all alarming examples of how ordinary members will be affected by this intrusive surveillance technology.

South Wales police have already used such cameras to monitor political dissent. In March 2018 the force used the cameras at a peaceful demonstration against an arms fair. Campaign group, Big Brother Watch, later established the force had put campaigners on the watchlist who were innocent and not wanted for any crime whatsoever. This is a worrying erosion of freedom, putting the UK on a path to a Chinese-style surveillance state.

And it could be worse. Imagine if the cameras misidentified a terror suspect. They could face on the spot arrest or worse without the legal rights we normally expect.

Furthermore, such false positives leave police forces swamped by masses of incorrect and misleading reports resulting time wasted searching for the proverbial needle in a haystack.

Live Facial Recognition cameras are just the tip of the iceberg. The last decade has seen a rapid expansion of the type of surveillance technology available to police forces. Automatic Number Plate Recognition cameras have become common place, the advances made in AI mean we are not far away from cameras which can lip read and recognise individuals simply by their gait.

The issue with all these technologies is they are used in a completely ungoverned space. There are ‘guidelines’ on their use. But nothing in the way of strict regulations with a statutory backing. The roll out of the technology has largely avoided parliamentary scrutiny of any sort.

The government must ensure the legislative framework keeps up with the rapid scale of technological advancement. The time to act on this is now while the technology is not as entrenched as other tools available to forces. If the government drags their feet on this, they will be left playing catch up.

In May, San Francisco – one of the most tech-friendly cities in the world – banned the use of live facial recognition technology because of the issues identified with it. An exception was made for federally controlled facilities at the airport and port, but the proactive stance taken by the city highlights the widespread concern on the roll out of these cameras.

A similar moratorium on the use of live facial recognition here in the UK would be the wisest way forward. This would provide the time and evidence to understand the risks and limitations of the technology to ensure facial recognition can only be used when the circumstances justify it and only in line with strict regulations.

We must not thoughtlessly let such technology become a ubiquitous part of modern life, eroding civil liberties in the UK for little gain and some risk.