Minister of Justice Andrew Little says police failed to get any of the necessary clearance before trialling controversial facial recognition software.
It follows RNZ revelations that police tested American company Clearview AI's program without consulting their own bosses or the Privacy Commissioner.
• NZ Police trialled facial recognition tech without clearance
• San Francisco bans police use of face recognition technology
• Border security turns focus to face-recognition technology
• Facial-recognition will get 'out of control' - experts
"I don't know how it came to be that a person thought that this was a good idea," Little said.
"It clearly wasn't endorsed, from the senior police hierarchy, and it clearly didn't get the endorsement from the [Police] Minister nor indeed from the wider cabinet ... that is a matter of concern."
Clearview, which is used by hundreds of law enforcement agencies in the United States and around the world, is effectively a searchable database lifted from the internet that can easily identify people once their images are uploaded.
The company was first approached by New Zealand police in January, and a short trial was later conducted, according to documents RNZ obtained under the Official Information Act. Police say they have decided not to use the product for the time being.
Clearview has sparked controversy for its opaque structure and its database of images ripped from social media sites - a violation of their terms of service. It claims to have been used to solve crimes from mailbox thefts to child sexual abuse, but has not been independently reviewed.
Little said he had received assurances from police that any further exploration of technology like Clearview would have "proper ministerial or cabinet approval".
He said he did not know what the trial was used for.
Police did not respond to any of RNZ's questions today on its relationship with Clearview.
The company pitched its product as a counter-terrorism solution, claiming proven success overseas.
However, Waikato University law professor Al Gillespie said the software would not have helped police in this area.
"The tools they've got are already good for the job if they're pointed in the right direction. You can look at something like Christchurch [terror attack] and they weren't pointed in the right direction. Giving that extra surveillance technology, I'm not sure they would have stopped that."
Gillespie said terrorists were often invisible until they acted, although limited surveillance around places like courts and airports could help.
Abdul Lateef Smith, a security contractor who has advised the Muslim community on security in the wake of the Christchurch attacks, said facial recognition technology could also be deployed at mosques.
However, he said it would only help if someone was already a person of interest.
The Green Party's police spokeswoman Golriz Ghahraman said terrorism had been used to justify eroding rights around the world, and New Zealand should not follow suit.
"This type of technology will have to be scrutinised very robustly."