Police use facial recognition to scan one million faces in WA trial leading to 122 arrests
2026-11-02
The Guardian
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Western Australia Police conducted a five-month trial of live facial recognition AI, scanning 1.38 million faces across 43 locations. The system generated 292 alerts, resulting in 122 arrests and 146 interactions with registered sex offenders. Nine incorrect alerts led to wrongful police engagement, raising concerns about privacy and potential rights violations.[AI generated]
Why's our monitor labelling this an incident or hazard?
The facial recognition system qualifies as an AI system because it performs real-time biometric identification and matching, a complex AI task. The system's use directly led to arrests and interactions with individuals, including some false alerts, which constitute realized harms such as potential violations of human rights (privacy and due process) and harm to individuals wrongly engaged. The concerns about algorithmic bias and privacy further support the classification as an AI Incident. The event is not merely a potential risk (hazard) or a response (complementary information), nor is it unrelated or solely beneficial use. Therefore, this event is best classified as an AI Incident.[AI generated]