Western Australia’s police have publicly disclosed comprehensive information about their facial recognition van trial, including deployment locations, arrest statistics, error rates, and privacy assessments. Meanwhile, Delhi Police operate similar vehicles with minimal public detail, raising questions about oversight and accountability.
- WA Police scanned 1.38 million faces, producing 292 alerts with 122 arrests
- Nine incorrect alerts accounted for about 3% of total alerts, no false arrests
- Delhi Police’s facial recognition van deployment lacks public transparency
What happened
Western Australia Police completed a five-month trial using vans equipped with facial recognition technology, deployed about 100 times across 43 locations in Perth and regional areas. The technology scanned approximately 1.38 million faces, generating 292 alerts that led to 122 arrests. The police force made public detailed reports including privacy impact assessments, equality impact assessments, and location breakdowns.
The trial utilized NEC’s NeoFace Watch software, technology comparable to that deployed by London’s Metropolitan Police. Non-targeted faces were pixelated in real time and not stored. The police commissioner emphasized the low error rate with nine incorrect alerts but acknowledged the ideal would be zero, and the vans were withdrawn from public deployment pending further evaluation.
Why it matters
Western Australia Police’s transparency contrasts sharply with Delhi Police’s lack of public disclosures despite operating facial recognition vans visibly labeled in public places. The WA trial allows public scrutiny of usage, effectiveness, and privacy protections, fostering informed debate about civil liberties and technological oversight.
Civil liberties advocates have highlighted concerns that even a small percentage of false alerts can adversely affect individuals and emphasized that reported false positive rates require validation through legal proceedings. The WA trial sets an important precedent for law enforcement transparency and the balancing of public safety and privacy rights.
What to watch next
As Western Australia reviews the trial outcomes and privacy considerations, stakeholders will look for decisions on whether to resume or expand use of facial recognition vans and the implementation of stronger regulatory frameworks. Input from privacy officials and civil society will be critical to shaping future policy and deployment safeguards.