What is police facial recognition and how do we stop it?

What is facial recognition?

Put simply, facial recognition is a form of technology that attempts to match a person’s face from a picture, video footage or live camera feed to a database of facial images.

Do the police use facial recognition?

Yes. Several UK police forces have used facial recognition technology since Leicestershire Police scanned thousands of people’s faces with it at Download Festival in 2015.

Police use has been spearheaded by the Metropolitan Police and South Wales Police. The forces have used the surveillance tech to scan hundreds of thousands of faces at protests, sporting events, concerts, Notting Hill Carnival, Remembrance Sunday, train stations, busy shopping streets, and even the seaside.

In 2022, it’s clear that the Met Police is ramping up its use of the tech.

How does police facial recognition work?

Police facial recognition works by measuring and ‘mapping’ a person’s unique facial features. These ‘maps’ are then converted into a numerical code to be matched against the codes of faces on secretive watch lists.

Up until now, the police have used what is known as ‘live’ facial recognition. This means that the act of scanning a person’s face and comparing to the watch list happens in real time.

It usually involves facial recognition cameras mounted on top of police vehicles stationed in crowded areas. The cameras scan the faces of everyone in range, and the software instantly compares them to the database.

Recently, South Wales Police announced that it was testing facial recognition on officers’ phones, so they could more easily scan people’s faces in the street.

As well as ‘live’ facial recognition, the Met has purchased software that allows it to carry out what is called ‘retroactive’ facial recognition. This is when faces in still images or previously captured video footage are scanned and compared to the watch list.

Who is on the watch list?

The police say the watch lists are made up of dangerous criminals and people wanted by the courts. There are usually thousands of people on a watch list each time a force uses facial recognition.

In 2020, we represented Ed Bridges against South Wales Police (SWP) in the world’s first legal challenge to the use of live facial recognition – more on this below (spoiler, we won).

During the case, we were able to see SWP documents that revealed that anyone could be on the watch list, whether they were wanted in connection with a crime or not. It also became clear that the images on the watch list could come from anywhere. The police could even take them from our social media accounts.

What’s the problem with police facial recognition?

The ‘maps’ that facial recognition makes of your face is unique to you. Much like a fingerprint, it is identifiable biometric data.

With ‘live’ police facial recognition, cameras scan everyone in sight, so this data is likely being snatched from you without your knowledge or consent. And this is certainly the case with ‘retroactive’ police use.

This is gross violation of your human rights.

Police officers have previously admitted to us that they just deploy live facial recognition in crowded areas to scan as many people as possible.

Retroactive facial recognition also turns every photo or video available to the police – including any you upload to social media – into a possible surveillance tool.

Does it work?

Lots has been said about the inaccuracies of the tech and how incorrect matches with the watch list have led to harmful police interactions.

Studies show that it particularly struggles to tell Black people apart, and has difficulty with women of any ethnicity.

But the idea that more accurate tech would lessen the problems with facial recognition is false. History shows that surveillance technology will always be used to monitor and harass people of colour. More accurate tech would only make this easier and discriminatory policing worse.

Is it lawful?

No.

When Liberty and campaigner Ed Bridges took South Wales Police to court for its use of live facial recognition, the Court said the force’s use of the tech was unlawful because it violated everyone’s human rights.

The Court also said that SWP hadn’t adequately taken account of the discriminatory impact of the tech – failing to meet its obligations under equality laws.

And by processing people’s unique biometric data, SWP also breached data protection laws.

You can read more about our legal victory here.

The Met Police has pushed ahead and continues to use live facial recognition after our win against SWP, but the Met’s use must also violate human rights, equality and data protection laws – and is therefore unlawful.

We also believe that retroactive facial recognition is similarly unlawful.

What should happen?

Especially after our court victory, there have been calls for Parliament to create laws governing police use of facial recognition.

However, there is no way of creating a law that solves the human rights or data protection issues, let alone discriminatory policing.

The safest thing to do is ban police from using dangerous facial recognition surveillance technology. More than 80,000 people have signed our petition. Add your name today.