A Māori technology ethicist says the government is allowing tech companies to use Māori as guinea pigs for facial recognition technology.
A Māori Reference Group at the Ministry of Social Development (MSD) said this was okay because the Identity Check technology would gradually improve.
An MSD report in September said it would inform its Māori and Pasifika Reference Groups "of the unconfirmed risk relating to racial bias, or the level to which it may exist, in DIA's technology".
The risk was "unknown" because no tests specifically for racial bias have been run here by the Department of Internal Affairs (DIA), which has been developing the system since 2019.
But Dr Karaitiana Taiuru disagreed that any racial bias was okay, and said members of the Māori Reference Group should front up and justify their thinking.
Speaking to Morning Report, Taiuru said it was not good enough that the technology was being rolled out without consultation with the Māori community.
There were "mountains" of international research and case studies which proved these systems were racist and would cause issues for any person of colour, such as Māori and Pasifika, he said.
"From what I'm reading and looking at, it looks like Māori beneficiaries will be used as the guinea pigs for an international system to recognise Māori and Pasifika faces and it almost feels like just because they are beneficiaries, they're being treated differently than non-beneficiaries."
Taiuru agreed that it could help people in being able to access benefits online instead of going into offices, but it was not known where data would be stored.
He said the technology was not likely to recognise faces of people of colour and they could only try a certain amount of times before being locked out and having to wait 72 hours to try again.
This could delay people getting help - and the system could falsely identify them as someone else.
Taiuru said if he were leading the roll-out, he would be engaging heavily with the privacy commissioner, and Māori communities to make sure they were aware of the "discrimination and racism facial recognition technology causes around the world".
He said solutions needed to be looked at to make sure such systems work in New Zealand in a way that benefited everyone.
In a statement, Privacy Commissioner Michael Webster said his office would begin consulting on new rules for biometrics - such as facial recognition technology.
"New Zealanders need to have trust and confidence in the use of biometrics by organisations and businesses.
"My office will issue a biometrics code exposure draft in early 2024 that we'll open for everyone to have their say on.
"Biometrics affects us all and I want to hear from the experts and other stakeholders we usually hear from, but also from the people going to the supermarket, or receiving marketing, who will have views on what parts of their personal information is collected and stored."
Webster said he was already aware Māori had concerns over bias and discrimination.
"We're taking a leadership position here because we need to develop ideas that are workable and effective but also take into account technological advancements.
"We have looked at the privacy risks related to biometrics, analysed what is happening with laws in other countries, and heard from local stakeholders. It has shown us that consulting on new rules specifically for biometrics is the right way forward."