The Government has passed a law that lets the Ministry of Social Development lean far more heavily on automated systems when it decides who gets a benefit and on what terms, after the Social Security (Modernisation) Amendment Bill cleared its third reading in Parliament on Friday.
The legislation, first flagged in Budget 2025, widens the Ministry’s existing power to make automated decisions about welfare. The Government says the change is about speed and consistency rather than cutting people off, and insists it does not alter who is eligible for support.
Social Development and Employment Minister Louise Upston framed the bill as a way to take routine paperwork off the desks of frontline staff. She said the goal was to “free up time for case managers to engage better with clients so they can access assistance and move from welfare to work”. The Ministry, she argued, would use automation only for straightforward, rules based decisions that do not need a person to weigh up competing factors.
Upston was at pains to stress that real people would stay in the loop where it counted. “The system will not replace human judgement where it is appropriate, and in many cases will make it easier for people to confirm their circumstances,” she said. The Government also says the bill “strengthens existing safeguards for the use of automated decision-making, including legislative requirements to manage bias, ensure transparency and maintain appropriate human oversight”, and that decisions will stay fair, accountable and open to challenge.
National’s Scott Simpson, speaking during the debate, said the current system left clients and taxpayers waiting too long for answers. “That’s not good enough for the clients of MSD, or taxpayers. This Bill fixes that,” he told the House, adding that “automated decision making will be used for simple, rules based decisions, and human judgement will remain where it is needed”. You can read more on the proposal in 1News coverage of the bill.
The Government’s coalition partners lined up behind the change. ACT MP Parmjeet Parmar pointed to the time savings for staff and said “nobody’s taking that human element out”. New Zealand First’s Jamie Arbuckle called the bill “a significant step towards a more efficient, modern welfare system”.
The Opposition was far less convinced, and much of the debate circled back to one word from across the Tasman. Several MPs raised Australia’s Robodebt scheme, an automated debt-recovery programme that wrongly told tens of thousands of welfare recipients they owed money to the government. A royal commission later found the scheme was unlawful, that it made people feel like criminals, and that it was linked to deaths. For critics, it is the cautionary tale that hangs over any move to let computers make calls about people’s incomes.
Green Party social development spokesperson Ricardo Menéndez March was the sharpest. He described the bill as a “carte blanche expansion to basically allow a robot, a machine, to have power of people’s lives” and reminded the House that “people died in Australia because of automated systems”. In a statement after the vote, he said “handing decisions about people’s ability to survive to a machine is a deliberate choice to make it harder for people to get the help they need”.
Labour’s Helen White worried about the people most likely to fall through the cracks of a more automated system. “You’re talking about the very group of people who are most disconnected,” she said, arguing that the New Zealanders who deal with MSD often need human contact rather than a faster algorithm. Te Pāti Māori’s Oriini Kaipara questioned whether the promised safeguards could really hold, warning that “technology isn’t neutral when the system itself is unequal”.
The Government has been clear about one point that often gets muddled in the headlines. The automated decision-making in the bill is not the generative artificial intelligence that powers chatbots like ChatGPT. It refers to rules based systems that apply settings already written into law, the kind of process that can, for example, confirm a payment when someone’s details match what the system expects. Ministers say the harder, judgement heavy decisions, and anything involving a sanction or a cut, will still pass through human hands.
Even so, the speed of the bill’s passage drew criticism. Some MPs argued it had moved through the House without enough consultation or scrutiny for a change that touches one of the most vulnerable groups of people the state deals with. The Government counters that the safeguards written into the law, and the limit to simple decisions, are exactly the guard rails that Australia lacked.
Where the line actually sits between a “simple” decision and one that needs a human will now fall to the Ministry to work out in practice. That detail matters, because it is the difference between an efficiency gain that frees up case managers and a system that quietly makes consequential calls about people’s money with very little oversight. You can follow the latest on the debate through RNZ’s political coverage.
The welfare debate is shaping up as one of the dividing lines heading into this year’s general election, with the major parties offering sharply different visions of how firm and how automated the safety net should be. If you want to see where each party stands before you vote, you can compare their positions using our voting tool.
What do you think? Should routine benefit decisions be handed to automated systems if it means faster service, or are the risks shown by Australia’s Robodebt too great to ignore? Share your view in the comments below.
This article was written by AI, briefed to report the facts, hopefully without some of the bias people bring to the job 🙂
Leave a Reply