Fears as force restarts using facial recognition technology on streets
FACIAL recognition technology is being used once again in South Wales despite a new warning that it has “serious implications for human rights and civil liberties” and “without sufficient safeguards, supervision and caution... could undermine a range of human rights, risk the fairness of trials and damage the rule of law”.
The controversial software was rolled out by South Wales Police in 2017. It maps faces in a crowd by measuring the distance between features, then compares results with a “watchlist” of images. This can include suspects, missing people and persons of interest.
In August 2020 the Court of Appeal ruled its use was unlawful after a legal challenge was brought by civil rights group Liberty and Ed Bridges from Cardiff. The court also found its use was proportionate interference with human rights as the benefits outweighed the impact on Mr Bridges. South Wales Police said it would not be appealing the findings.
South Wales Police has begun using it once more and it was used for five hours from noon on Saturday, March 19.
The force says it is part of an “independent evaluation” in response to the Court of Appeal judgement. In that trial, which had a watchlist of specific suspects, three positive identifications were made resulting in two arrests – one of the suspects was for an assault on a police custody officer and the other for failing to appear at court for a public order offence.
A third man wanted for breaching a supervision order ran off from the area on seeing police officers.
The force failed to answer how many future trials it will use or who is monitoring the results of that trial.
In a statement released on March 14, Police and Crime Commissioner Alun Michael said he has “close scrutiny of the operational decisions on the introduction of technology and subjects each new step to independent oversight and scrutiny because of the ethical and social concerns that have been expressed over the use of facial recognition technology”.
He added: “People want to know that members of the public who have done nothing wrong are not being subjected to inappropriate surveillance, and that their privacy and anonymity will be respected and protected. However, people also want us to keep them safe and to use the technology to apprehend people who have committed serious offences and take them off the streets. As a result of the robust systems of scrutiny and challenge in place in South Wales, I can assure the public that we are getting that balance right. We are committed to protecting human rights as well as keeping the public safe.”
A report released yesterday by the House of Lords Justice and Home Affairs Committee said that there are dangers in the technology.
It said: “Without sufficient safeguards, supervision and caution, advanced technologies used in the justice system in England and Wales could undermine a range of human rights, risk the fairness of trials and damage the rule of law. Facial recognition is the best known but other technologies are in use and more are being introduced. Development is moving fast and controls have not kept up. The committee acknowledges the benefits: preventing crime, increasing efficiency and generating new insights that feed into the criminal justice system.
“However it is concerning that there is no mandatory training for the users of AI technologies, such as facial recognition, particularly given their potential impact on people’s lives. Meanwhile users can be deferential (‘the computer must be right’) rather than critical. The committee is clear that ultimately decisions should always be made by humans.
“There are risks of exacerbating discrimination. The report highlights serious concerns about the dangers of human bias contained in original data being reflected, and further embedded, in algorithmic outcomes. The committee heard about dubious selling practices and claims made as to products’ effectiveness which are often untested and unproven.”
Discrimination in relation to facial discrimination was discussed in the Senedd at plenary on March 22. Plaid Cymru leader Adam Price said the first use of the technology since the Court of Appeal’s judgement was on the weekend of a large anti-racism rally in Cardiff.
Addressing the First Minister he said: “According to the UK Government’s own biometrics and forensics ethics group the lack of representation of ethnic minority faces in the training data on which the technology used by the police is based means it is more likely to identify innocent black people as criminals. This will exacerbate the racial disproportionality in rates of detention that you, yourself, have acknowledged. In Scotland the use of this technology is banned for this reason. We lack the power to do so currently in Wales but will you at least support the prohibition of its use on publicly-owned land like the Senedd steps where the rally was held on Sunday?”
Mark Drakeford said he was “very well aware of the concerns” about face recognition technology.
He said: “I think those concerns deserve to be taken very seriously. I know that my colleague Jane Hutt has had an opportunity to discuss this and allied matters with the lead PCC for Wales, Dafydd Llywelyn, and we will continue to make sure those concerns are properly represented to PCCS and indeed to chief constables where it’s an operational matter.”
The College of Policing published a guide for officers in England and Wales on Tuesday to make sure their use of live facial recognition technology is “legal and ethical”.
People want to know that members of the public who have done nothing wrong are not being subjected to inappropriate surveillance, and that their privacy and anonymity will be respected and protected. However, people also want us to keep them safe and to use the technology to apprehend people who have committed serious offences and take them off the streets
Police and Crime Commissioner Alun Michael