Auckland, 21 November 2025: New research from Kantar reveals only 4% of New Zealanders feel well-informed about how artificial intelligence (AI) is being used by government agencies, with 90% knowing little or simply being aware of its existence without understanding the details. This is a key finding in Kantar’s new study ‘New Zealanders’ Views of AI Use in the Public Sector’ released today. It provides the most comprehensive snapshot yet of public sentiment, trust, and expectations around AI use in government services.

When asked what describes their current understanding of how AI is being used in the public sector, half (49%) claim to ‘know a little’ and a significant two in five (41%) state ‘I’ve heard of it but don’t know how it’s being used’. Of the two in five (39%) of people who say they have ‘personally interacted with an AI in the public service’, it is mostly through chatbots. Notably, the sector with the highest self-rated knowledge is health where 18% of respondents reveal awareness, mainly for notetaking and receptionist support.

While AI is now in widespread use across the public sector in New Zealand for everything from data analysis and automation to customer service; public understanding remains limited, says Kantar New Zealand Group Account Director – Public Sector, Jay Carlsen.”Trust and comfort with AI in the public sector are evenly split with just over half of New Zealanders (55%) saying they are ‘comfortable with government agencies using AI,’ but the remaining 45% remain uneasy. Trust is fragile, with only 5% expressing strong confidence in responsible government use of AI. People are concerned about data security, privacy, and the accuracy of AI-driven decisions and they seek robust regulation.”

The data shows that half want government agencies to have independent oversight or regulation of their AI systems (51%) and for staff to be trained on ethical and responsible use (48%), while 46% want to ensure its only used in low-risk or non-critical services. Notably 17% overall and 36% of Māori want agencies to follow Māori sovereign data principles when handling iwi or hapū data.

“The majority of New Zealanders support the creation of an independent government body to oversee AI use,” adds Carlsen, “yet awareness of existing regulations is low, and few believe current safeguards are adequate.”

Despite these concerns, the study reveals a cautious consumer optimism about the potential of AI to improve public services.

“New Zealanders value AI for its ability to enhance efficiency in straightforward tasks, but they insist on human oversight for complex or sensitive decisions,” says Carlsen.

“Two-thirds want the option to interact with a human rather than AI (66%), and many would prefer to opt out of AI-driven services altogether (53%). The greatest perceived value for AI lies in areas such as transport, digital infrastructure, and cybersecurity, while there is considerable scepticism about its

use in justice, social services, and policy development, where human judgement and empathy are seen as essential.”

The overarching sentiments reveal a clear need for building public confidence in AI, which will require clear communication, robust regulation, and transparency about how AI is used and how decisions are made.

“Agencies must ensure ethical use, provide options for human interaction, and involve the public in the design and rollout of transformative AI services,” adds Carlsen.

“There is strong opposition to AI handling personal or confidential data without human oversight, especially in decisions that could significantly affect people’s lives. New Zealanders recognise the potential of AI to improve public services, but their support is conditional on trust, transparency, and the assurance that human values remain at the centre of decision-making. Our research highlights the need for government agencies to engage openly with the public, strengthen regulatory frameworks, and ensure that the benefits of AI are realised without compromising fairness or accountability.”