‘Minimum ethical and legal standards’ urged for UK facial recognition
A University of Cambridge report argues that all UK law enforcement should be required to use a new audit to ensure responsible use of biometric recognition. A school researcher applied the audit to three police departments to see what would happen. All three failed.
Among other policy recommendations made in the 151-page paper, police use of facial recognition in public spaces should be banned.
The conclusion applies to all types of face biometrics for identification, including live, retrospective and mobile phone applications, according to author Evani Radiya-Dixit, who researches facial recognition policymaking.
Radiya-Dixit, a visiting fellow at the university’s Minderoo Centre for Technology and Democracy, tested her audit on three police facial recognition deployments in England and Wales.
They were the South Wales Police operational biometrics trial deployments between 2017 and 2019 (which were eventually ruled unlawful by the UK court of appeal), the Metropolitan Police Service’s trial of live facial recognition between 2016 and 2019 and the South Wales Police’s trial of mobile phone facial recognition between 2021 and 2022.
“We found that all three deployments failed to meet the (proposed) minimum ethical and legal standards for the governance of facial recognition technology,” the report says.
They failed to comply because the trials were “very broad in scope” and may have “infringed upon privacy rights.” And the face biometrics systems deployed on those occasions “were not transparently evaluated for bias in the technology or discrimination in its usage.”
The deployments did not ensure that there was a reliable human in the decision-making loop. Nor were there clear redress measures for people harmed by the use of facial recognition, the paper found.
Finally, the research indicated that these deployments lacked regular oversight from an independent ethics committee and the public.
“For example, the ethics body overseeing South Wales Police’s trials had no independent experts in human rights or data protection, based on the available meeting notes. South Wales Police also did not consult the public or civil society for feedback before their trials,” reads the document.
There “have been improvements in how police use facial recognition, but more work needs to be done.”
The author makes three recommendations to protect people from algorithm errors and misuse.
The first calls for regulators, civil society groups and researchers to use the proposed audit to scrutinize police use of facial recognition. They should also evaluate the use of biometric technologies used in other contexts and regions. The third recommendation calls for a ban on police use of facial recognition in public spaces.
“To protect human rights and improve accountability in how technology is used, we must ask what values we want to embed in technology and also move from high-level values and principles into practice,” concludes the report.
It comes days after the Scottish Biometrics Commissioner published its first report arguing Parliament should have confidence in how biometric technologies are being used for policing and criminal justice purposes in that country.
biometric identification | biometrics | ethics | facial recognition | law enforcement | police | regulation | surveillance | UK