Your staff adopted AI years before any party bothered with a policy

New Zealand businesses did not wait for permission. Most large firms already use artificial intelligence, many staff use it daily, and the productivity case is well rehearsed. Only now, with voters about to head to the polls, has AI become a live election issue. RNZ’s analysis notes that after months of near-silence, Labour has a policy, Opportunity has sketched one, and ACT’s David Seymour has finally said something substantive.

That is a lag measured in years, and business is the one paying for it.

Firms moved while officials studied

The adoption numbers are not new. A 2024 Datacom survey, cited in MBIE’s 2025 Cabinet paper, found 67% of larger NZ businesses used some form of AI, up from 48% in 2023.

The policy machinery ran on a different clock. In June 2024 Cabinet’s Economic Policy Committee opted for a proportionate, risk-based approach using existing law rather than a standalone AI Act. That was a sensible call. But a Treasury OIA release from August 2025 showed officials were still working through AI’s productivity implications more than a year later. The national strategy, when it landed in mid-2025, was light-touch and focused on SME uptake, with no new regulation attached.

Light-touch is defensible. Slow is not.

The real gap is in small business

The headline adoption figure hides the problem. In 2025 MBIE reported that 68% of NZ SMEs had no plans to evaluate or invest in AI at all, against 38% of Australian SMEs. That same page put New Zealand 40th of 188 countries on government AI readiness.

For an economy built on small firms, that is the productivity story. Large corporates have IT budgets and legal teams to manage the risk. A 12-person engineering consultancy does not, and uncertainty over liability, privacy and data rules is exactly what keeps it on the sidelines. Treasury’s own 2024 economic analysis leaned on Productivity Commission and OECD work showing technology diffusion, not invention, is where small economies win or lose.

Three parties, three instincts

Labour has the most developed package. Its policy announcement promises an Office of AI inside DPMC with a responsible minister, an independent online safety regulator, copyright rules for creators, data standards aligned with Australia, renewable-energy conditions on data centres, and an AI small business fund. Some of it is useful. A new office and a new regulator is also a lot of new bureaucracy for a country that already decided existing law could carry most of the load.

Seymour sits at the other end. He told RNZ the public service is “too slow” to adopt AI and sees no reason it should not draft Cabinet papers. His broader position is that New Zealand should be a “fast follower” of overseas rules rather than pretending it can shape global regulation. That is the most realistic thing any politician has said on the subject.

The Greens want a one-year moratorium on new AI data centres, according to Newsroom’s expert roundup. RNZ’s analysis flags the obvious flaw: blocking local capacity pushes more New Zealand data offshore, the opposite of sovereignty. It would also cap investment in one of the few infrastructure classes offshore capital actively wants to build here.

Copy the homework, but do it now

The experts and Seymour land closer together than either might admit. Academics including Dr Andrew Lensen and Dr Cassandra Mudgway have signed an open letter, covered by Newsroom, calling for bipartisan, risk-based regulation and a national oversight body. In August, University of Auckland associate professor Gehan Gunasekara argued for aligning with the EU or Australia so compliance is simple for both big tech and local firms. Lensen warned of a “soft power risk” to New Zealand’s trade reputation if it has no comparable framework.

For exporters selling into Europe and Australia, that matters more than any domestic ethics debate. Mirroring Canberra’s settings is cheap, low-overreach and saves firms building two compliance regimes.

What business actually needs is short: clarity on who is liable when an AI tool errs in a regulated industry, procurement rules for AI in government contracts, data residency settings firms can plan around, and practical help for the SMEs still sitting out.

Whoever forms the next government should treat that as a first-100-days list, not a discussion document. Business has already spent two years answering the AI question on its own. The risk now is that politicians, having finally noticed, spend the next three years debating it.

Sources

• RNZ: AI has finally entered the election campaign arena (2026-10-03)
• RNZ: ACT leader David Seymour doesn’t see why AI shouldn’t write government Cabinet papers (2026-10-02)
• RNZ: Labour promises to set up AI regulator, copyright rules if elected (2026-09-09)
• Newsroom: What political parties should consider for their AI election policies (2026-10-01)
• The Spinoff: If AI is about to upend everything, where’s all the AI policy? (2026-08-05)
• Treasury: Analytical Note 24/06 – The Impact of Artificial Intelligence, an economic analysis (2024-07)
• Treasury: Official Information Act Response 20250500 – Future of AI (2025-08-13)