What's facial recognition ?
People often think facial recognition is just the next generation of CCTV, but it’s a far more intrusive form of surveillance than that. CCTV just takes pictures, facial recognition takes measurements of our face – the distance between the eyes, length of the nose, shape of the face, etc – it’s essentially the next generation of finger printing.
Forced into police line ups
Police live facial recognition surveillance vans have recently invaded our public spaces in Croydon. They target marginalised communities – scanning their faces to steal their biometric data without their consent, treating them like suspects forced into a digital line up against a police watchlist every time they walk near the van.
Inaccurate
Independent studies based on the police’s own data accumulated over 8 years shows that 85% of the time the system gets it wrong with an innocent person, not on any police watchlist, accused of being a criminal (2016-2023).
Racial Bias
The technology is known to particularly discriminate against people of colour and women. The police’s own research shows that at certain settings 22 out of 33 people falsely identified were black. When the system was deployed in 2017 at the Notting Hill Carnival – the largest black African Caribbean event in the UK, 96 alerts were triggered but only one of them was confirmed correct – that’s a 99% failure rate (95/96 = 98.96%), which was still an improvement from the previous year which had a 100% failure rate!
Police Dishonestly
Note that these raw figures were obtained through freedom of information requests, the police’s own massaged statistics are wildly dishonest. For the 99% error rate at Notting Hill Carnival they claim just a 0.09% error rate as they disingenuously compare against, not the total number of people they stopped, but rather the 101,000 people who happened to walk past the van ( 95/101000 = 0.094%).
When the parliamentary committee on Science, Innovation and Technology questioned Lindsey Chiswick, Director of Intelligence and National Lead on Facial Recognition for the Metropolitan Police, about the high failure rate she simply denied it: “I don’t recognise those figures”, saying there were “zero” false matches! When she was given a specific example of a police documented false match she replied “I don’t have details of that specific incident with me” (24 May 2023).
Assumption of Guilt
Another problem is that the police end up believing their own lies. In a recent interview Inspector Crystal Govers (Ealing Met LFR deployment lead) when asked about the concerns of Ethnic minorities on the biases and errors in LFR , she replied “I’m going to go out on a limb. It is a flawless system, I believe” (9th July 2024). The presumption of innocence does not exist against a “flawless system”.
Those wrongly picked by the “flawless system” are assumed guilty until they can prove their own innocence. The burden of proof against a “flawless system” can be incredibly high. Sometimes a clear ID document is not enough, and they are asked to submit to more intrusive biometrics like fingerprints (Oxford Circus, Jan 2022).
The police claim on notices near their LFR vans that “there is no legal requirement for you to walk through the LFR system” but in reality, they have had plain cloths police stationed near the notices and they have stopped people who have decided not to walk past their LFR vans. Also, they have stopped people who have partially covered their face whilst walking past the vans and forced them to submit to their face being scanned, even giving one person a £90 fixed penalty notice for his indignation at the ordeal (Romford, May 2019). People are assumed guilty for not wanting to be scanned by police cameras.
Accuracy Can't Solve Systemic Racism
Baroness Louise Casey’s recent report (March 2023) reconfirmed that 25 years on from the Macpherson report nothing has changed – the Metropolitan Police Service is still institutionally racist. It’s important to understand that whilst the technology has been shown to be inaccurate and biased against people of colour, even an accurate facial recognition system will still produce racist outcomes in society, as in the hands of an institutionally racist police force the technology will inevitably be disproportionately used against communities of colour and Muslims – two groups who have historically been subjected to over-policing. For example, tasers work accurately on all races, but they are used disproportionately on people of colour (3x more, 2015). We oppose facial recognition not because of its bias, but because it’s a tool of oppression in the hands of an institutionally racist police force.
We oppose facial recognition not because of its bias, but because it’s a tool of oppression in the hands of an institutionally racist police force.
Andy George, President of the National Black Police Association just this month (July 2024) disclosed that the Met is “setting up LFR to disproportionately impact the Black community in London”, deploying it into “almost singular black communities rather than other places”.
Also, if the watchlists used by the LFR system for its targets is disproportionately made up of people of colour then even a 100% accurate system will still disproportionately target people of colour. The Met refuses to reveal the demographic breakdown of the watchlists it handpicks for targeting in facial recognition deployments. Despite being repeatedly asked three times by the parliamentary committee on Science, Innovation and Technology “Can we get a demographic breakdown of the watchlist?”, Lindsey Chiswick, Director of Intelligence at the Metropolitan Police refused, saying “there is no policing reason” (24 May 2023).
Campaigners from Big Brother Watch who attend LFR deployments have observed that the vast majority of people stopped by the Met since 2020 following an LFR alert have been people of colour, mostly black people.
Other biometric systems like mobile fingerprint scanners are known to be disproportionately used against people of colour – a black person is 4 times more likely to be stopped and scanned than a white person (2020-2022).
To get a better understanding of the dangers of facial recognition, especially for ethnic minorities, its worth looking at it in the wider framework of racialised surveillance.
Racialised Surveillance
Racialised surveillance has its origins in colonialism, and slavery. It was a tool for controlling enslaved people from Africa whilst they were on ships and on plantations. Biometric policies, such as fingerprinting, were tools pioneered within the British Empire to identify civilians and prisoners from the colonies. And this hasn’t stopped, racialised surveillance is still used today as a tool to control people of colour.
Today most countries in Africa are being forced by the World Bank through its Sustainable Development Goals to implement digital IDs, it starts by it being a requirement in order to access basic services inside your country, and then is used to track you when you leave your country. The UNHCR is also complicit, refugees have to hand over their biometrics in order to access food, these are then shared with governments so they can track and control refugees. Hungry citizens who have had to use UNHCR food stations have also been snarled in this net, as once scanned they are put into a UN refugee database without their knowledge and are suddenly at risk of being deported from their own country (eg Kenya).
Racialised Surveillance In The UK
Racialised surveillance is also extensively used in the UK to control its minority communities.
Section 60 Stop & Search
Today we see the targeted surveillance of the black community using Section 60 suspicion-less Stop and Search powers where a black person is 12 times more likely to be stopped and searched without reasonable suspicion. More than 60 children a week are being strip-searched by police in England and Wales, with those who are black, Asian or mixed race significantly more likely to be targeted.
CCTV
A study on ‘CCTV and the social structuring of surveillance’ conducted by the Centre for Criminology and Criminal Justice at the University of Hull revealed that black people were twice as likely (68%) to be surveilled for ‘no obvious reason’ than white people (35%) by CCTV operators.
Prevent
We see attempts to control Muslim communities by systemic surveillance under the controversial counter-terrorism strategy, known as ‘Prevent’. Prevent duty was introduced specifically to target the Muslim community, with its budget being allocated based on a local area’s Muslim population size determined from the census (Amnesty).
It turns teachers, doctors and social care professionals into spies for the state – over 5m public service employees have been trained to report any signs of ‘radicalisation’. In the health sector Muslims are 8x more likely to be referred to Prevent than non-Muslims.
Most of its victims are children – more than half of all Prevent referrals are children under 15. Recent Prevent training for teachers reveals children wearing Palestinian insignia, or a girl wearing hijab when her mother doesn’t, should be reported as signs of radicalisation. No age is too young – even a 3 year old toddler has been reported to Prevent!
Prevent operates in the ‘pre-crime space’ on the dangerous assumption that a criminal act can be prevented before an individual ever plans, or even intends it. To do so, it must ignore key legal principles including the assumption of innocence and the need for a solid evidential basis for suspicion. Amnesty has described Prevent as “thought police” saying it’s “fundamentally incompatible” with the UK’s international human rights obligations.
We also see attempts to ban of the Muslim face veil, the niqab, under security pretexts (Imperial College, Birmingham Metropolitan College).
Surveilling Communities -Project Champion
Another example of the racialised surveillance of a whole minority community was in 2010 when Project Champion was rolled out – a counter-terrorism surveillance strategy which involved deploying over 200 cameras targeted at Muslim suburbs of Birmingham (unprecedented number at the time). 72 cameras were hidden and over 100 were equipped with Automatic Number Plate Recognition systems to record car registrations as they moved around the area.
The police installed the cameras under the Safer Birmingham Project, lying to the community that it had received £3 million from the home office to improve community safety, to combat car crime and anti-social behaviour. In reality the cash had come from the Terrorism and Allied Matters fund to surveil the Muslim population of Washeood Heath and Sparbrook wards.
The cameras were finally removed after mass mobilisation of the community demanding they be taken down. Even when the decision was finally taken to remove them, and it was announced that in the interim plastic bags would be put over the cameras, the police refused to put plastic bags over the hidden cameras as that would reveal their locations, hoping to keep them whilst promising to remove them.
Surveilling Migrants
The racist surveillance of refugees is yet another example of racialised surveillance. The UK and foreign governments have shared facial biometric surveillance data of migrants with reach other, and on the basis of the data received, the UK government has rejected asylum claims.
For those whose asylum claims are being considered, their biometric data is forcefully taken from them, and their imposed immigration bail conditions insist they must report to a home office reporting centre every week. With only 13 reporting centres in the whole country the journey can take many hours. The home office’s Croydon reporting centre is one of these. Many migrants have been arbitrarily abducted during reporting, taken to be imprisoned indefinitely at a deportation centre.
Others are forced to wear electronic GPS tagging surveillance devices clamped to their ankles, which monitor them 24/7. We have seen some refugees that despite being surveilled by ankle devices are still being required to regularly attend reporting centres.
The Guardian has reported that as part of the Home Office’s Satellite Tracking Service, migrants who have been convicted of a criminal offence will be required to scan their faces up to five times a day using smartwatches installed with facial recognition technology. Their locations will be tracked “24/7, allowing trail monitoring data to be recorded”.
Lucie Audibert, legal officer for Privacy International, said “No other country in Europe has deployed this dehumanising and invasive technology against migrants.”
Dr Monish Bhatia (Criminology at Birkbeck, University of London), said: “Electronic monitoring is an intrusive technology of control. Some individuals develop symptoms of anxiety, depression, suicide ideation and overall deterioration of mental health. The Home Office is still not clear how long individuals will remain on monitoring. They have not provided any evidence to show why electronic monitoring is necessary or demonstrated that tags make individuals comply with immigration rules better. What we need is humane, non-degrading, community-based solutions.”
Surveilling The Hungry
In Croydon we are blessed with a diverse population, but within that, due to systemic racism, black and Asian children are disproportionately affected by poverty. Black and Asian children are 2.3x more likely living in poverty than white children. 42% of our guests at the Resistance Kitchen are Asian, and another 37% are Black. At our community kitchen we have a strict policy of serving all in need without asking any personal questions, or requiring a referral or registration or any other method of surveilling our guests.
But we are shocked to learn that some food banks in London have started to trade biometric data for food. A mobile app is used to scan the faces of vulnerable people to enable them to receive ‘free’ food from participating stores. The food banks say it reduces their work load and prevents fraud. 96% of our guests at Resistance Kitchen are food insecure – one in six hasn’t eaten for at least a whole day in the last month – we would question if such desperate people have agency to give informed consent for their most personal data being taken from them.
This particular example, whilst not of surveillance by the state or its security apparatus, is nevertheless relevant in showing how private enterprises are also using racialised surveillance of vulnerable communities, often the distinction is blurred.
We have to understand that the use of live facial recognition by the police today is just the latest tool in this matrix of racialised surveillance unleashed on marginalised communities in order to control them.
Facial Recognition Unlawful
There is no absolute legal basis for the use of live facial recognition let alone any specific legislation regulating its use, and in 2020 the High Court on appeal held that the police’s use of facial recognition was unlawful – breaches privacy rights, data protection laws and equality laws. Following this the Equality and Human Rights Commission has called for its suspension.
Last October 2023 this call was echoed by a cross-party parliamentary group of 65 MPs and Lords, backed by 31 human rights groups including Amnesty, Liberty, and the Race Equality Foundation.
There are currently two legal actions underway against the use of facial recognition by the Met police and by private shop security company Facewatch – both victims are people of colour who have been misidentified as criminals (May 2024). Shaun Thompson, working in a Croydon community youth outreach group which works to take knives off the streets, was pulled up and held for thirty minutes by Met Police after he was misidentified as a criminal by facial recognition on his way home near London Bridge. He was only let go after handing over a copy of his passport. Thompson said: “Facial recognition is like stop and search on steroids and doesn’t make communities any safer. It needs to be stopped.”
“Facial recognition is like stop and search on steroids and doesn’t make communities any safer. It needs to be stopped.”
Accelerated Rollout
Despite these serious shortcomings, the Minister for Policing – local MP Chris Philp (now ex-Minister) has written to police forces urging them to double their use of facial recognition, to unleashing it on our community. The Met Commissioner has increased use of live facial recognition by a 1,000% in London. Bankrolled by the Mayor of London, London is now the third most heavily surveilled city in the world – more than Beijing (PrivacySavvy 2021).
Croydon Mayor Jason Perry defended the use of facial recognition in Croydon by regurgitating the old mantra ‘if you’ve got nothing to hide you’ve got nothing to fear’. Whistleblower and anti-surveillance advocate Edward Snowden replied to this argument saying “Arguing that you don’t care about the right to privacy because you have nothing to hide is no different than saying you don’t care about free speech because you have nothing to say.”
Fearmongering
Politicians and the police cower the community through fear and misinformation, into submitting to this most contentious surveillance of our community.
Croydon Police Chief Superintendent Andy Brittain justified its use by stoking fear in the community saying “I think it’s a vital tool for the Met to use to keep London safe. We’ve had some real crime problems in the past here and the community have reached out to us saying, ‘I want to feel safe here, I want to be able to come out, go shopping and not worry about crime.”
The perceived fears are then used to support the draconian surveillance of our community. The Chair of Croydon’s Safer Neighbourhood Board claimed “Women, feeling afraid to walk through our own town centre when they need to shop, that supersedes the concerns around ‘walking identity cards’. For us here in Croydon, it’s really important that we’re kept safe. Children and people have died here because of the level of criminality so it’s important for us to take a stand… if it’s going to aid the borough commander in extra policing and keeping Croydon safe, I’m all for it.”
In reality LFR takes police officers away from their beat rather than add extra policing. A civil liberties observer attending a LFR deployment noted “Recently, I went to Croydon to watch how police were using live facial recognition in the area. In the five minute walk from the station down to the deployment, I witnessed two men put on balaclavas and rob a corner shop in broad daylight. There wasn’t a police officer in sight. When I got to the deployment, there were 15 officers standing around, gazing at tablets waiting for supposedly wanted criminals to walk past their cameras and the “facial recognition in use” signs, essentially handing themselves in for arrest.” She added “It’s a terrible use of taxpayers’ money. But the worst cost is our liberty. High crime rates make us less free but so too does Orwellian policing. Live facial recognition has no place in London.” (May 2024)
The observed single LFR deployment of 15 officers (that doesn’t include officers in the van, or back end support, or preparation) for 5 to 8 hours amounts to around 100 hours that could be spent on the beat safeguarding the community from violent crime.
The police go on insisting LFR is to help fight violent crime in our city centres. Chief Superintendent Sean Wilson, Borough Commander for the Met’s West Area sees LFR surveillance as a “deterrent” against violent crimes, specifically knife crime (July 2024).
But this contradicts the police’s own research – 29 different studies all show that CCTV has no impact on violent crime – zero deterrence, and 33 studies all showing that it does not reduce crime in city centres (College of Policing report, Oct 2021).
The police insist that it’s to help fight violent crime in our city centres. But this contradicts the police’s own research – 29 different studies all show that CCTV has no impact on violent crime and 33 studies all showing that it does not reduce crime in city centres (College of Policing report, Oct 2021).
Croydon has now become a hotspot for police facial recognition vans.
Lessons From Project Champion
We need to learn the lessons from Project Champion. Firstly, the police lie – they may introduce LFR as a benign ‘safe neighbourhoods scheme’ telling us it’s to fight knife crime or violence against women… anything to get their LFR surveillance cameras up and running. But once they are normalised in our community, they will become permanent, and they can then be used for any nefarious purpose without consulting the community like we saw in Birmingham.
Secondly, if we come together as a community we can defeat this pervasive surveillance of our community as they did in Birmingham.
Resistance
If live facial recognition isn’t stopped now then those clumsy vans will soon be replaced by permanent installations in our city centres. And the future looks even bleaker – there is nothing technically stopping all of the UK’s 6 million CCTV cameras, that the police can tap into, from feeding their facial recognition system. When connected to our driving licence and passport facial database it would create a complete surveillance state where every citizen is identified and monitored from the moment they leave their home.
Many cities around the world, including San Francisco, have woken up to the dangers of facial recognition surveillance systems and have banned them. Several London local councils, including Haringey, Newham, and Islington, have passed motions calling on the Met police not to use facial recognition in their boroughs.
Dystopian future
In calling for the ban, Newham councillor Areeq Chowdhury, who is also Head of Policy at The Royal Society whose founding mission is to promote the use of science for the benefit of humanity, warned fellow councillors that “Dystopian Futures do not occur overnight, they are built slowly piece by piece with each piece disguised as a minor technical adjustment to our existing way of life and then before you know it you look around and realize that science fiction mirrors reality and the rights we once took for granted are now impossible to exercise”.
“Dystopian Futures do not occur overnight, they are built slowly piece by piece with each piece disguised as a minor technical adjustment to our existing way of life and then before you know it… the rights we once took for granted are now impossible to exercise”.
Used Againt Non-Criminals
Already we are seeing live facial recognition surveillance being used by the police in non-criminal contexts. It has been deployed against peaceful protesters (eg Cardiff Arms Fair in March 2018, Remembrance Sunday 2017) with watchlists consisting not of wanted criminals, but of people the state doesn’t want attending these events. Facial recognition being used as a tool to control society, to prevent people exercising their political rights, engaging Articles 10 (freedom of expression) and 11 [freedom of assembly] of the Human Rights Act.
Predictive Policing
Police watchlists have also included innocent people who have not broken the law but who the police think may break the law in the future – yes like the dystopian sci-fi movie Minority Report where the ‘PreCrime’ police stop crimes before they take place by arresting people who have done nothing wrong but are predicted to commit crime in the future by ‘PreCogs’ who can foresee crimes.
The guidance from the College of Policing on who can be put on a watchlist includes individuals who haven’t committed any offence, but you “suspect… is about to commit an offence” and for good measure you can also add their “close associates” (includes family?) who you do not suspect of wrong doing in the past, the present or the future, onto the watchlist (March 2022).
Such predictive policing systems hardwire historic racist policing into present day policing and law enforcement practices.
This is also very prevalent in the private sector where shop facial recognition systems are used to predict consumer behaviour, and have watchlists of people who have not been convicted of a crime but who the shop predict may shoplift or be antisocial, and are undesirable (eg Facewatch).
The Met have also been caught sharing watchlist photos with private security companies after initially denying it (Granary Square 2019).
Often the private sector, where regulations are more lax, is used as an extension of the states surveillance apparatus. Private companies are given free rein to collect consumer biometric data and surveillance video which then the state has unhindered access to it. Amazon’s innocuous Ring door bell is an example – since the video (and audio) is stored in the cloud belonging to the company, the company gives the police access without need for a warrant, or the owners consent. In return the police recommends the product and are even handing them out free to household in areas they want to surveil. One in five households in the UK has a video door bell (2022).
Resistance Kitchen
From Croydon city centre the vans have now alarmingly been rolled out to Thornton Heath, just a few streets from our kitchen.
Allowing these invasive new powers to be given to a police force that has been repeatedly found to be institutionally racist, and seems incapable of reform, is unacceptable – we have to resist.
The Resistance Kitchen, along with other community groups are launching a campaign to stop this assault on the basic human rights of our community – stop police facial recognition surveillance now!
Lets work together to stop this dystopian future, lets get rid of police live facial recognition, starting from Croydon!
*the original article was updated on 23/7/2024