Face Off: The Lawless Growth of Facial Recognition in UK Policing

Face Off: The Lawless Growth of Facial Recognition in UK Policing is the May 2018 campaigning report published by Big Brother Watch that opened the British civil-society campaign against police use of automated facial recognition surveillance. Built on more than 50 Freedom of Information requests to UK police forces and on observed deployments at the 2017 Notting Hill Carnival and the 2017 Remembrance Sunday commemorations, the report is the British anchor of the European civil-society case against live facial recognition and the foundational evidence base for the Stop Facial Recognition campaign that Big Brother Watch has run continuously since.

Findings

The report's headline finding is that the Metropolitan Police's automated facial recognition deployments had been wrong in more than 98% of matches since the technology entered operational use at the 2016 Notting Hill Carnival: 102 false positives against 2 correct identifications, neither of which was a criminal suspect. The corresponding figure for South Wales Police — the other British force then deploying automated facial recognition at scale — was a 91%-wrong-match rate: 2,451 false positives from 2,685 total matches, with 15 arrests representing 0.05% of matches. The report also documented Leicestershire Police's 2015 facial-recognition scan of 90,000 festival attendees at the Download Festival, the retention of innocent people's biometric photos for thirty days at the Metropolitan Police and twelve months at South Wales Police, and a Police National Database holding 19 million custody images, of which 12.5 million were biometrically searchable — including images of innocent people whose retention had been ruled unlawful by the High Court in 2012 but never expunged.

The report's framing of the deployments — captured in its subtitle — was that the policing programme had grown without primary legislation, an independent regulator, parliamentary debate, or democratic mandate, and that the combination of high false-match rates, indefinite retention of innocent people's biometric data, and demonstrable race-disproportionate impact in the police-tested populations placed it beyond the threshold at which incremental safeguards could be a sufficient response.

Recommendations

The report set out three named recommendations that anchored the Stop Facial Recognition campaign's substantive position: that UK public authorities immediately stop using automated facial recognition software with surveillance cameras; that unconvicted individuals' images be automatically removed from the Police National Database in line with the 2012 High Court ruling; and that any future deployment of automated facial-recognition technology by public authorities be brought under independent oversight through an established statutory board. The first of those — the immediate-stop demand — became the through-line of every subsequent Big Brother Watch facial-recognition product, from the July 2018 legal challenge against the Metropolitan Police and the Home Secretary, through the May 2023 Biometric Britain report, to the 6 October 2023 Joint Statement on UK Live Facial Recognition signed by 65 parliamentarians and 31 rights and race-equality organisations.

Parliamentary launch and 15-NGO coalition

The report was launched in Parliament in May 2018, with a follow-on civil-society and cross-party press event in June 2018 at which Big Brother Watch was joined by a 15-organisation civil-society coalition — Big Brother Watch alongside Article 19, defenddigitalme, the Football Supporters' Federation, Index on Censorship, the Institute of Race Relations, Liberty, The Monitoring Group, Netpol, the Open Rights Group, the Police Action Lawyers Group, the Race Equality Foundation, Race On The Agenda, the Runnymede Trust, and Tottenham Rights. The coalition assembled at the launch is the corpus's first recorded British civil-society coalition convened on a single AI-and-surveillance demand, and structurally prefigures the broader 31-organisation coalition the same organisation would assemble five years later for the October 2023 joint statement. The named parliamentary contributors to the report — Layla Moran MP, Nick Hurd MP as Minister of State for Policing, Baroness Williams as Minister of State for Countering Extremism, Norman Lamb MP as Chair of the Science and Technology Committee, the Lord Bishop of St Albans, and Baroness Jones of Moulsecoombe — bracket the early-2018 parliamentary debate that the report shifted onto the question of automated facial recognition's legal basis. Baroness Jones of Moulsecoombe went on to be the named lead claimant alongside Big Brother Watch in the subsequent 14 June 2018 legal challenge against the Metropolitan Police and the Home Secretary, brought by Rosa Curling at Leigh Day with Stephen Cragg QC and Adam Straw of Doughty Street Chambers as counsel and crowdfunded via Crowdjustice on Articles 8, 10 and 11 ECHR grounds.

Guest contributors and international framing

The report's body chapters carry named guest contributions from four civil-society and academic registers outside the UK domestic privacy cluster: Samir Jeraj of the Race Equality Foundation on the race-equality dimension of facial-recognition deployment in policing; Jay Stanley of the American Civil Liberties Union on the US comparative picture of police facial-recognition adoption and challenge; Clare Garvie of the Georgetown Law Center on Privacy and Technology on the United States Police-and-Face-Recognition database analysis that her centre's work had established; and Jennifer Lynch of the Electronic Frontier Foundation on the broader American civil-liberties critique of biometric mass surveillance. The cluster of contributors places Big Brother Watch's domestic policing analysis inside the wider international civil-society case against facial-recognition deployment in policing, and is the first formal point at which the corpus's UK and US biometric-mass-surveillance critique registers are jointly anchored on a single British campaigning artefact.

Posture in the movement

Face Off sits in the make-AI-good-movement record as the foundational publication-side anchor of the British civil-society campaign against AI-driven biometric mass surveillance — the report that converted the question of automated facial recognition's legal basis in the United Kingdom into a public-facing campaign demand and that supplied the 98%-wrong-match headline figure that the legal challenge, parliamentary engagement, and later coalition work all draw on. As a publication-side artefact the report stands as the first of three named Big Brother Watch reports that bracket the corpus's coverage of the UK automated-facial-recognition fight: Face Off (May 2018) on the lawful-basis case at the start of operational deployment, Biometric Britain (May 2023) on the operational-evidence case after five years of deployment, and the Joint Statement on UK Live Facial Recognition (October 2023) as the cross-party-plus-coalition consolidation of the immediate-stop demand. The report's specific contribution — establishing the 98%-wrong-match figure as the working evidential premise of the British case against automated facial recognition — places it alongside Gender Shades (2018) and Stochastic Parrots (2021) in the corpus's foundational evidence-side anchors of the algorithmic-accountability publication record, distinguished from those by its sustained engagement with the policing-and-surveillance application register rather than the commercial-AI-systems application register the other two foreground.