The Policing Minister has written to police chiefs to encourage the use of facial recognition technology and AI surveillance ahead of a safety summit this week.
Chris Philp urged the doubled usage of facial recognition software and cameras, coupled with AI for tracking offending. This letter was published ahead of a global AI safety summit to be held in Bletchley Park later this week.
AI facial recognition technology works by biometrically scanning faces and measuring features in order to map out an individualised vector for identification. Minister Philp says this vector would allow authorities to ‘stay one step ahead of criminals.
Proponents of this facial recognition technology point to the recent successful captures of three wanted suspects. However,. Fears regarding the increased usage’s impact on discrimination, privacy and human rights issues have prompted calls for cessation from a cross-party group of 65 peers and MPs, and have been echoed by organisations including Big Brother Watch, Amnesty International and Race Equality Watch. Former Brexit Secretary and member of cross-party group David Davis has called live facial recognition ‘a suspicion-less mass surveillance tool that has no place in Britain’.
This technology has previously been criticised by an independent team of academic researchers from University of Cambridge for its failure to meet “minimum ethical and legal standards”. This builds on a history of previous issues with police retention of personal images, with a watchdog referring to the lack of checks on balances on police access to sensitive data as an ‘omni-surveillance society.
However, the Home Office have dismissed these concerns, pointing to stringent guidelines when using facial data to counter concerns. These guidelines include visible notices in spaces where it is in use and automatic data deletion if no facial match is found. They cite optimization of police time and resources and dismiss concerns over discrimination and privacy, arguing that it would only be used where necessary and proportionate.
