Risks of bringing in facial recognition technology for policing ‘too high’, civil liberties groups to tell politicians
An Oireachtas committee is set to hear that the Garda has come across approximately 22,000 hours of footage from the Dublin riots, as they consider the benefits and drawbacks of introducing facial recognition technology (FRT) for policing.
Civil liberties groups will be informing the politicians that the risks of introducing FRT are too high and that it is not a solution that can guarantee success.
On the other hand, Garda Commissioner Drew Harris is expected to tell the committee that the manual processing by Garda personnel is becoming unfeasible and ineffective.
Mr Harris believes that every major criminal investigation now involves processing digital evidence, often images or footage from seized devices or CCTV. He points out that there has been an “explosion in the volume of digital footage as evidence”, and cites the example of the November 23rd riots. Initially, gardaí had reported that they were looking at 6,000 hours of footage, but TDs and Senators will hear on Tuesday that the footage from the riots runs to 22,000 hours.
Individual murder investigations can have “upwards of 50,000 hours of footage” and “seized devices can have over a million images of child sexual abuse”, the Oireachtas Committee on Justice will be told.
The Data Protection Commission (DPC) has stated that before introducing new technology, a Data Protection Impact Assessment (DPIA) must be conducted. The Garda has to process available footage to identify or exclude suspects, according to two separate judgments from the Court of Appeal.
The court found that the balance between a suspect’s right to privacy and the human rights of the victim must be maintained.
The Garda’s use of artificial intelligence technology has created understandable public concern and confusion. However, the Garda has reassured the public that there is no question of autonomous machine decision-making. All decisions that could potentially impact individuals are made by identifiable and accountable personnel. Modern biometric identification systems are reliable, and there must be proportionate safeguards in place for specific use cases.
