More than 1.3 million people have had their faces scanned by Western Australia Police during a contentious five-month trial of real-time mobile surveillance technology that culminated in 122 arrests.
The trial, which concluded on Saturday, saw specially equipped surveillance vans deployed 100 times across 43 regional and metropolitan locations, The Guardian Australia reported. According to official figures released by the state police force, biometric cameras captured 1,380,882 individual facial scans as members of the public moved through monitored public thoroughfares.
Officers using the system received alerts when the software identified potential matches against police databases of wanted individuals and persons of interest. Over the course of the operation, police carried out 122 arrests and engaged in compliance checks with 146 registered sex offenders.
Fleet of mobile surveillance vans deployed across state
The operation represents one of the largest public deployments of live facial recognition technology by an Australian police force to date. Purpose-built vehicles fitted with high-resolution cameras were stationed at high-footfall sites, capturing crowd footage and comparing facial measurements against law enforcement watchlists in seconds.
While police operational commands view the capability as an effective tool to detect suspects who might otherwise evade routine patrols, the sheer volume of members of the public captured by the cameras has reignited debates around passive surveillance. Unlike traditional security footage reviewed retrospectively following an alleged offence, real-time scanning indiscriminately processes the biological identifiers of everyone within the camera's optical path.
False matches and civil liberties concerns
Despite the arrests, official data confirmed the system was not infallible, recording nine false alerts during the five-month operational window. In those instances, officers were alerted to individuals who matched biometric parameters on the database but were entirely innocent of any wrongdoing.
Civil liberties advocates and legal researchers have routinely cautioned that even a nominal margin of algorithmic error carries significant human consequences, including unjustified stops and searches on public streets. Privacy organisations have repeatedly raised concerns that deploying automated biometric dragnets without explicit public consent sets an alarming precedent for state surveillance, eroding the basic expectation of anonymity in open community spaces.
Scrutiny over regulatory oversight and biometric data
The trial comes amid an evolving national dialogue regarding the regulation of automated decision-making and artificial intelligence in domestic policing. Unlike jurisdictions overseas, such as the European Union where real-time remote biometric identification in public spaces is heavily restricted, Australia lacks a unified, legally binding statutory framework governing the deployment of facial recognition by state police.
Legal experts have emphasised that current privacy statutes have struggled to keep pace with rapid developments in mass biometric scanning. Key questions remain over how long non-matching data is retained, how security watchlists are curated, and what independent mechanisms exist to review operational errors.
With law enforcement agencies across the country closely tracking Western Australia's findings, civil society groups are urging state and federal governments to enact strict legislative safeguards before temporary trials become permanent features of public life.
