(file photo) Photo: 123rf
New Zealand police have confirmed their staff have accessed not one, but two controversial facial recognition websites using police-issued devices.
The second site is called Facecheck.ID, which police said staff connected to 274 times last year.
Morning Report on Thursday reported on an internal audit into the use of the first website, PimEyes, which found 22 staff had accessed it.
Both sites scan hundreds of millions of images to find matches for uploaded facial photos, but were not authorised for police use and had sparked official complaints and controversy in the UK and Australia.
The information about Facecheck.ID was in the same police OIA response about PimEyes. Dated December 2023, it was obtained by an Australian news organisation.
Police released the information on Thursday - six days after RNZ requested it.
The new information showed their staff made 393 connections to PimEyes last year, until they blocked the site on 8 May 2023.
"Police believe that staff accessing Facecheck.ID were simply looking into what the tool was, and not using it in any material way," they said in a statement to RNZ on Thursday
"There was no evidence of breach of regulations or any harm."
One site "visit" could trigger multiple connection requests, so it would be incorrect to say there were 274 individual visits to the site, incorrect to say this constituted "use" of it, and incorrect "to imply there was misuse by police staff or use of the sites for police business", police said.
However, police were "comfortable" with RNZ reporting there were 274 connections via Facecheck.ID, they said.
Facecheck.ID had been blocked, though no audit was done of it by police, "most likely because it was blocked when we became aware of it", they said.
"Given that we had already distributed the policy banning the use of open source FRT [facial recognition tech] systems, and there was no evidence to suggest misuse, no further action was deemed necessary."
