The trial will use Operator-Initiated Facial Recognition (OIFR) technology, which allows officers to scan and identify people immediately during stops, according to multiple reports. Around 100 devices will be available for the pilot, which is scheduled to last six months. Mayor Sadiq Khan stated that the pilot would give the Met an opportunity to test and evaluate the capability of handheld facial recognition. He added that the technology is intended to ensure officers can check and verify details of individuals stopped instead of arresting them and taking them to a police station. Khan emphasized that the technology may not be rolled out if results aren't as beneficial as expected.
Oversight for the trial will be provided by the Mayor's Office for Policing and Crime (MOPAC) and the London Policing Ethics Panel to ensure its use is right and proportionate, multiple reports indicate. The Equality and Human Rights Commission has called for a new independent oversight body to regulate the use of facial recognition technology in the UK, according to the commission. Specific safeguards and regulations for the handheld trial to prevent misuse or bias have not been detailed publicly, leaving questions about how ethical concerns will be addressed in practice.
Called on the Met to immediately halt all LFR use until proper safeguards and regulations are in place.
The Met has been using facial recognition technology since 2020, deploying cameras on vans and in fixed locations, according to multiple reports. Face scanning has already been deployed by police with cameras on vans and in fixed locations including in Croydon, Manchester and South Wales, research shows. Retrospective facial recognition systems are widely in use across the UK, research indicates. The Met said there have been over 1,400 Live Facial Recognition (LFR) arrests since the technology was first introduced, with more than 1,000 people charged or cautioned.
The Met's increasing use of LFR technology has sparked concern from civil liberties campaigners and some London Assembly Members, multiple reports indicate. The Met is facing a High Court challenge from campaigners over live facial recognition technology, according to multiple reports. Campaigners say live facial recognition scans faces in public spaces without sufficient safeguards and risks unfair or discriminatory use, according to groups. The Met describes live facial recognition as a tool to help fight crime.
Londoners deserve transparency regarding the expansion of police powers with handheld facial recognition.
Youth worker Shaun Thompson and Silkie Carlo of Big Brother Watch spearheaded the legal action against live facial recognition, multiple reports indicate. Shaun Thompson was previously misidentified by the live facial recognition system, according to multiple reports. Thompson and Carlo voiced concerns that live facial recognition could be used arbitrarily or in a discriminatory manner. Lawyers representing the pair argued that facial recognition data is similar to a DNA profile and that permanent installations would make it impossible for Londoners to move freely without biometric data being captured.
Scotland Yard defended the legal challenge, telling the court that the live facial recognition policy was lawful. Lord Justice Holgate and Mrs Justice Farbey said the policy provides adequate indication of when live facial recognition will be used and enables people to foresee consequences reasonably, according to multiple reports. The judges said Shaun Thompson and Silkie Carlo's human rights have not been breached by the live facial recognition policy, multiple reports indicate.
The Met's website explicitly says they do not use handheld facial recognition technology.
The government previously defended plans to expand the use of facial recognition across England and Wales, increasing vans from 10 to 50 and making them available to all forces, according to multiple reports. The exact criteria and locations for deploying the handheld facial recognition devices during the pilot program have not been disclosed, raising questions about where and how the technology will be tested.
A contradiction exists regarding whether the Metropolitan Police currently uses operator-initiated facial recognition technology. The Met's website states it 'does not presently use the so-called operator initiated facial recognition', research shows. However, the Met is starting a pilot program for operator-initiated facial recognition technology. Operator-initiated facial recognition is already in use by South Wales police, where officers run NEC's NeoFace algorithm on their smartphones, research indicates.
The biggest breakthrough for catching criminals since DNA matching.
Facial recognition errors have occurred in practice, such as when police arrested a man for a burglary in a city 100 miles away that he had never visited after software confused him with another person of south Asian heritage, research shows. The outcomes and findings from previous uses of facial recognition technology by the Met, such as error rates or effectiveness in crime reduction, remain unclear, as detailed public reports are limited.
Separately, the Metropolitan Police is considering using AI to help identify victims of online child sexual abuse and categorise imagery by severity, according to multiple reports. The Met investigated more than 5,400 child sexual abuse offences over the past year, with more than 1,300 children requiring safeguarding, multiple reports indicate. The Met believes AI could help identify potential new victims earlier and shorten intervention time.
There is a danger that these technologies can be inaccurate and falsely identify people.
The Met said investigations into online child sexual abuse and exploitation cases are increasing year on year, and it manages over 12% of cases nationally. Officers currently manually review child sexual abuse material to link victims to known cases or identify unknown victims, according to multiple reports. Child sexual abuse content is graded according to severity across categories A, B and C, multiple reports indicate.
The Met is in discussions with technology companies about how AI tools could assist with identification without subjecting staff to graphic content, according to multiple reports. The Met is considering technology that would allow officers to review and risk-assess 641,000 messages in about 35 minutes, multiple reports indicate. How the AI tools for child sexual abuse investigations will be implemented and what specific technologies are being discussed with companies have not been specified, leaving implementation details uncertain.
The data shows that there are racial disparities for false positive identification, causing human rights infringements and distress to those affected.
The Met said any use of AI would operate within strict legal, ethical and safeguarding frameworks, with specialist officers retaining decision-making responsibility. The use of AI in policing has attracted controversy, particularly over live facial recognition, according to multiple reports. Deputy Commissioner Matt Jukes stated that AI could significantly reduce the time officers and staff are exposed to distressing material.
Additional context reveals that the pilot was announced on Thursday when Sadiq Khan responded to questioning from an opposition politician, research shows. The Met also signed a £490,000 three-month contract with Palantir to try to detect rogue officers based on their wider conduct, research indicates. According to The Guardian - Main UK, opposition politician Sarah Jones described facial recognition as the biggest breakthrough for catching criminals since DNA matching.
Critics have raised alarms about the expansion. According to Evening Standard - News, Green Party London Assembly Member Zoë Garbett described the need for the Met to immediately halt all LFR use until proper safeguards and regulations are in place. She also described that Londoners deserve transparency regarding the expansion of police powers with handheld facial recognition. According to The Guardian - Main UK, Garbett described the handheld trial as an alarming change, noting that officers will be able to literally walk up and scan people's faces on the device.
Human rights advocates warn of risks. According to The Guardian - Main UK, human rights advocate Mary-Ann Stephenson described that there is a danger that these technologies can be inaccurate and falsely identify people. She also described that the data shows racial disparities for false positive identification, causing human rights infringements and distress to those affected. Details of the independent oversight body proposed by the Equality and Human Rights Commission to regulate facial recognition technology have not been elaborated, leaving its structure and authority undefined.