On 20th February, the Metropolitan Police deployed facial recognition technology for the second time in London, this time at Oxford Circus. The cameras scan the faces of people walking by and compare them against a known ‘watchlist’ so that the police can approach them if they are flagged. The Met claims that the technology creates a false alert only one in 1000 times and that “this technology helps to keep Londoners safe.” However, the only independent review carried out on behalf of the Met from the University of Essex found that the technology resulted in only eight correct matches out of 42, across six separate trials. There are two clear arguments to be resolved: the first being who is correct and the second asking whether the use of this technology is proportionate in the fight against crime, or a concerning power grab by the police.
The basic argument for facial recognition is clear: that it makes it much harder for suspected criminals to evade police capture. Traditional CCTV technology has proved important in the fight against crime, however it has its limitations. Although there is CCTV coverage of virtually all of London, with the highest number of CCTV cameras per person of any city outside China, the police do not have sufficient resources to check all of this footage. In comparison, facial recognition harnesses the power of technology to quickly flag suspects without human intervention. In the context of a rising violent crime and murder rate in London, this argument is likely to be attractive to those in search of an answer.
However, it is also important to consider the nuances of how this technology works. By taking the picture of everyone who walks by and comparing it to a database, it is essentially performing a search on you. How different is this from being stopped by a police officer and asked to show identification to compare against a known database? It is maybe faster and less inconvenient, but no less intrusive. It is interesting to note that the UK is one of the first Western countries to trial this technology, despite our longstanding queasiness with the concept of “your papers, please?” Only 10 years ago, the newly elected coalition government repealed an ID card law which had been extensively campaigned against, and the current majority Conservative government is now permitting the roll out of this potentially far more intrusive technology.
Plans for facial recognition technology extend far beyond the UK. The technology has already been deployed in China for some time, and is even being used to police jay-walking with offenders identified and flagged to the police. Closer to home, leaked European Union documents call for legislation to create and interlink national police facial identification databases across member states. The rapid spread of the technology in authoritarian countries is not a surprise, as policy can be implemented quickly with limited public or legal scrutiny. More disturbing is the creeping extension of powers to use facial recognition to authorities in democratic countries, often with little public debate and no specific legislation to regulate these powers. Indeed, the Met Police have justified their recent use of facial recognition through a court ruling regarding South Wales Police’s use of the technology, which occurred in the markedly different context of football games and is still subject to appeal. The High Court ruled that the use of the technology met the conditions set out in previous data protection legislation, namely the Data Protection Act 2018. While this is the current legal interpretation of the issue, a poll in September 2019 said that 55% of people think the government should limit police use of facial recognition to specific circumstances. In this context, the use of this ruling to justify the implementation of a powerful new technology against considerable pushback is problematic, and shows a disregard to UK norms of policing by consent and democratic oversight.
On 24th February, the Met Police commissioner, Cressida Dick, defended the technology and called critics “ill-informed.” She claimed that it is “for the critics to justify to victims of crimes why police shouldn’t use tech lawfully and proportionately to catch criminals.” The problem with this argument is that it assumes that the all the police have done is to implement a relatively benign new technology. In reality, it is for the police to justify to the public why further erosion of their rights and suspicion-less identification is suddenly necessary to combat crime. In London, from a starting point of a reputation tarnished by years of failings and misconduct, and the only independent evidence thoroughly discrediting the technology, this argument will need to be strong.