New Zealand businesses have adopted artificial intelligence (AI) faster than almost any technology before it. A Deloitte Access Economics study found 82 percent of businesses now use AI in some form. Over the same period, national labour productivity fell 0.7 percent. Almost everyone has switched AI on. Far fewer have changed how they work because of it.
That gap between adoption and value is not a technology problem. It is a leadership problem. Closing it is precisely the job of a Chief Information Officer (CIO), or for mid-sized organisations that cannot justify the full-time headcount, a virtual CIO (vCIO). If nobody in your business owns that job, the job still exists. It is simply not being done.
Here is what your CIO or vCIO should be doing for you right now.
Turning AI noise into an investment thesis
Every vendor you buy from is now an AI vendor. Every renewal comes with new AI features, new pricing tiers and new promises. Your CIO’s first responsibility is to filter that noise against your business strategy, not the vendor’s roadmap.
That means a clear position on where AI genuinely moves the needle for your organisation, where it is a distraction, and where it introduces risk you are not being paid to take. It means sequencing investment so you fund two or three initiatives that matter rather than twenty experiments that do not. AI’s share of technology spend in New Zealand is forecast to rise from 29 percent to 34 percent over the next two years. Someone needs to make sure your share of that spend is deliberate.
Governing AI use before it governs you
The Employers and Manufacturers Association (EMA) found 83 percent of small and medium-sized enterprises (SMEs) are already using AI. Only 13 percent have an AI policy. That is not an adoption gap. That is a governance vacuum.
Your people are already pasting customer data, contracts and financials into AI tools. Without a policy, you have no visibility of which tools, what data, or what obligations you are breaching. Your CIO should have already delivered an AI use policy your staff can actually follow, an approved tool list, and a process for evaluating new AI services before they touch your data. This is weeks of work, not months, and it is far cheaper than explaining a privacy breach to your customers and the Privacy Commissioner.
Getting your data ready to be useful
AI is only as good as the data you point it at. Most mid-sized organisations have a decade of files nobody has classified, permissions nobody has reviewed, and systems that do not talk to each other. Deploy an AI assistant across that estate and it will cheerfully surface every salary spreadsheet and board paper your permissions failed to protect.
Your CIO should be driving data classification, access reviews and information architecture as a precondition to AI deployment, not an afterthought. Unglamorous work, but it is the difference between AI that compounds value and AI that compounds risk.
Redesigning work, not just adding tools
This is why national productivity went backwards while adoption soared. Most businesses bought AI features inside tools they already owned and kept working exactly as before. The firms seeing real returns did something different. Deloitte’s analysis found the average AI-adopting SME generated roughly NZD 400,000 more in revenue in FY25 than a comparable non-adopter, with large firms showing substantially greater gains again. That is a revenue gap, not a cost saving, and Deloitte presents it as an early estimate of the relationship between AI use and firm performance rather than a guaranteed payoff.
The difference is process redesign. Your CIO should be sitting with your operational leaders, mapping how work actually flows, and rebuilding processes around what AI now makes possible. Licences without redesign is how you join the 82 percent who adopted AI and the majority who got nothing for it.
Building capability across your people
The EMA survey found 43 percent of businesses not using AI cite lack of expertise as the main barrier, and workforce readiness lags well behind adoption even among users. Your CIO owns closing that gap: structured training, identifying internal champions, and setting realistic expectations about what AI can and cannot do. A tool your people do not trust or understand delivers nothing, whatever the licence costs.
Holding the line on security and cost
AI disruption does not suspend the fundamentals. Cyber risk is rising as attackers adopt the same tools your business does, and AI spend has a habit of creeping across cost centres where nobody totals it up. Your CIO should still be delivering the basics: cyber resilience appropriate to your risk profile, vendor and contract management, and a technology budget the board can trust. AI ambition built on insecure foundations is a liability, not a strategy.
Measuring outcomes, not activity
Finally, your CIO should report AI progress in business terms. Not tools deployed or licences purchased, but hours returned to the business, error rates reduced, revenue protected or created. If your technology leadership cannot tell you what your AI spend returned last quarter, you do not have an AI strategy. You have an AI subscription.
The questions to ask this month
If you have a CIO, ask them three things.
- Where is our AI investment concentrated and why there?
- What policy governs how our people use AI today?
- What did our technology spend return to the business last quarter?
If nobody in your organisation can answer those questions, the role is vacant whether or not it appears on your organisation chart. A vCIO gives mid-sized organisations senior technology leadership at a fraction of a full-time executive’s cost, for exactly the moments like this one, when the technology landscape moves faster than your business can safely follow on its own.
Unisphere Solutions provides vCIO services to organisations across New Zealand. If any of the questions above made you uncomfortable, that is worth a conversation. Get in touch.