How the trial worked

Under the trial's guidelines, cameras mounted on or positioned near marked police vehicles scanned faces in designated public areas, matching them against an alert list drawn from police databases.

Police said the safeguards worked as follows:

• •Images of people who weren't on the list were pixelated in real time and deleted immediately, without being stored
• •Deployments took place across 43 regional and metropolitan locations in Western Australia
• •The scanning ran over five months and ended on a Saturday

Arrests, alerts and the Perth Royal Show

Police figures recorded 122 arrests and nine incorrect alerts from the facial recognition WA police trial. Alerts also led officers to engage with 146 registered sex offenders and to locate 10 people for a welfare check.

Three operations at the Perth Royal Show accounted for a significant share of the activity:

• •223,749 faces scanned
• •15 alerts generated
• •Three arrests
• •Seven interactions with registered sex offenders

Police defend the 0.00065% error rate

WA police put the error rate at 0.00065%, based on nine false alerts against the total of about 1.38 million scans.

At a Sunday media conference, commissioner Col Blanch said of the false alerts: "No one was arrested. No one made a complaint. Each one did result in an interaction with that person".

He said: "It's a very, very, very low number. It's not zero. We'd love it to be zero."

Blanch also conceded the technology was "not perfect" and that authorities "knew this going into this trial". He said the trial focused on using modern technology responsibly to identify people who presented a risk while most people went about their day unaffected.

Critics question the numbers and the privacy safeguards

Civil Liberties Australia chief executive Bill Rowlings said WA police were "severely gilding the lily" by claiming the incorrect alert rate was 0.00065%.

Rowlings argued the rate should be judged by convictions rather than arrests, saying: "If half the 122 arrested … are not convicted, the 'incorrect' rate will be 50%". He also said police had performed the almost 1.4 million scans "with no reasonable suspicion whatsoever".

When the trial was announced, the Office of the Information Commissioner WA said it was not involved in implementing it and warned of "collective privacy harms".

Experts and critics additionally warned the technology could disproportionately affect marginalised groups, noting facial recognition algorithms have historically had higher false-match rates for people with darker skin tones.

The surveillance vans have now been removed from public deployment while WA Police conduct an internal evaluation of performance and community impacts, including a review of safeguards.