UK immigration enforcement is planning another trial of live facial recognition at Holyhead, a key maritime connection between Ireland and the UK in Wales.
Travellers arriving in the UK from Ireland will be scanned and compared against a database of offenders and people of interest this week, The Irish Times reports.
The third “proof-of-concept pilot” follows a pair carried out in November, when Home Office says 7,500 faces were scanned, with no incorrect alerts and one arrest made. UK intelligence reports show individuals are using the Common Travel Area, which minimizes border controls between Britain and Ireland, to circumvent immigration controls.
“Immigration Enforcement (IE) have identified UK ports where intelligence indicates that individuals are either returning in breach of a Deportation Order or are using Common Travel Area routings to avoid formal scrutiny and detection,” according to a report titled “Immigration Enforcement – Live Facial Recognition Legal Mandate.”
The 12-page report describes the legal considerations and risk mitigation measures in place for such deployments. The document sets out the policy’s alignment with UK immigration, rights and data protection regulations.
It describes the watchlist as composed of “individuals subject to extant Deportation Orders or people who are wanted on warrant in relation to an immigration-related criminal offence.”
Civil liberties advocates warn that the trial imposes checks the Common Travel Area is supposed to avoid, and that they also embed surveillance infrastructure into the port that could be repurposed in the future.
Adopting AI and minimizing bias
The trial comes as part of a much broader embrace of facial recognition and other AI tools in law enforcement.
National Crime Agency Director of Threat Leadership Alex Murray told The Guardian this week that the new national police AI center will recognize the bias risks and work to minimize them.
“If you talk about live facial recognition or predictive policing, there will be bias, and you need to get in the data scientists and the data engineers to clean the data, to train the model appropriately, and then to test it,” said Murray.
“There is no point releasing something to policing that has bias in it that’s not recognised, and everything should be done to minimise it to a level where it can be understood and mitigated.”
The publication cites the discovery of demographic disparities in a legacy retrospective facial recognition algorithm from a December report as showing the need for stronger safeguards. The software had been updated, but Home Office had declined to use the new version, opting for one released just as the biometrics industry began grappling with the issue of uneven algorithm performance between people from different demographics.
Article Topics
biometric bias | biometrics | border security | demographic fairness | facial recognition | immigration | law enforcement | live facial recognition