Two years ago, we published our first NZ enterprise AI landscape report and found an ecosystem that was "highly interested but underinvested." The picture in 2025 is fundamentally different. AI has moved from boardroom conversation to operational reality, but the distribution of that progress is uneven, and the gap between leaders and laggards is now a chasm.
What You Need to Know
- NZ enterprise AI adoption has doubled since 2023. Approximately 72% of larger NZ businesses now report using AI in some operational capacity, up from 48% in 2023. But "using AI" spans a wide range, from Copilot licences to production AI platforms.
- The top tier has pulled away. Roughly 15% of NZ enterprises now have multiple AI capabilities in production on shared infrastructure. These organisations are seeing compound returns and accelerating. The remaining 85% are falling further behind.
- AI investment is concentrating, not spreading. Total NZ enterprise AI spend has increased, but it's concentrated among fewer, larger initiatives. The era of scattered $50K pilots is ending. Organisations are investing $200K-$1M+ in platform-level AI programmes.
- The talent constraint has shifted. The bottleneck is no longer AI specialists. It's AI-literate business leaders who can scope, govern, and operationalise AI. NZ has more data scientists than it has leaders who know what to do with them.
- Government and regulation have caught up. The NZ Government AI Framework (published mid-2025) provides clear guidance for public sector AI use. Private sector governance is maturing in response, though formal AI regulation remains principles-based rather than prescriptive.
72%
of larger NZ businesses using AI in operational capacity in 2025
Source: NZTech, AI Readiness in Aotearoa 2025
$7.2B
estimated AI productivity opportunity for the NZ economy by 2030
Source: Microsoft New Zealand, AI Economic Impact Study (updated), 2025
The State of the Market: Five Key Findings
1. The Three-Tier Model Has Consolidated
In our 2023 report, we described three tiers: Explorers, Experimenters, and Operators. Two years later, the tiers have shifted.
| Tier | 2023 | 2025 | Movement |
|---|---|---|---|
| Explorers (aware, no structured initiative) | 60-70% | 28% | Significant reduction |
| Builders (1-3 capabilities, establishing foundations) | 20-25% | 57% | Major growth |
| Operators (production platform, compound value) | 5-10% | 15% | Steady growth |
NZ Enterprise AI Maturity Shift: 2023 vs 2025
Source: NZTech, AI Readiness in Aotearoa 2025; RIVER Group
The middle tier, which we've renamed "Builders," has absorbed most of the former Explorers. These organisations have moved from conversation to action. But moving from Builder to Operator is proving harder than moving from Explorer to Builder. The leap requires infrastructure investment, governance maturity, and organisational change that many are still working through.
2. Investment Levels Are Maturing
The average AI investment profile for NZ enterprises has changed significantly:
| Investment Category | 2023 Typical | 2025 Typical | Change |
|---|---|---|---|
| Discovery and strategy | $20-50K | $30-80K | +50% |
| First production capability | $80-150K | $120-250K | +60% |
| Annual operational spend | $50-100K | $120-300K | +150% |
| Foundation/platform investment | Rare | $200K-800K | New category |
Annual AI Operational Spend Growth (NZ enterprises)
Source: NZTech, AI Readiness in Aotearoa 2025; RIVER Group
The most notable shift: "Foundation/platform investment" has emerged as a distinct budget category. In 2023, almost no NZ enterprises were investing in shared AI infrastructure. In 2025, the Operators are spending $200-800K on platforms that serve multiple capabilities, and seeing compound returns that justify the investment.
3. Governance Has Gone From Optional to Expected
In 2023, fewer than 15% of NZ enterprises had any formal AI governance. In 2025:
NZ Enterprises With Formal AI Governance (%)
Source: NZTech, AI Readiness in Aotearoa 2023-2025
52%
of NZ enterprises with formal AI governance frameworks in 2025
Source: NZTech, AI Readiness in Aotearoa 2025
The NZ Government AI Framework, published in mid-2025, accelerated this shift, particularly for organisations that work with government. Private sector governance has matured in parallel, driven by board-level awareness and the practical reality that governance accelerates deployment rather than blocking it.
Key governance trends:
- Risk-tiered approaches are standard. Organisations classify AI use cases by risk level and apply proportionate controls.
- Data sovereignty is explicit. NZ organisations are increasingly specifying where data is processed and stored, driven by both customer expectations and emerging regulation.
- Audit logging is becoming infrastructure. The Operators have automated, continuous monitoring. The Builders are implementing logging as they deploy.
4. The Talent Picture Has Shifted
The AI talent market in NZ has evolved in ways we didn't fully anticipate:
What's improved: The supply of AI technical talent has increased. University programmes have expanded. International recruitment has stabilised. Cloud provider training programmes have upskilled thousands. The raw "data scientist shortage" story from 2023 is less acute.
What's worsened: The bottleneck has moved upstream. The scarcest resource in NZ enterprise AI is now AI-literate business leadership: executives, product owners, and operations managers who understand AI well enough to scope opportunities, set governance, and drive adoption. Technical talent without business-aligned leadership produces technically impressive capabilities that nobody uses.
What's emerging: A new role, the AI product owner, is appearing in NZ enterprises. This person sits between business and technical teams, translating business needs into AI requirements and managing the AI capability lifecycle. Organisations with this role are deploying faster and achieving higher adoption.
5. Sector Leaders Are Emerging
AI maturity varies significantly by sector in NZ:
| Sector | Maturity | Leaders | Focus Areas |
|---|---|---|---|
| Financial services | High | Banks, insurers | Claims, fraud, compliance, customer service |
| Government | Medium-High | Central agencies | Document processing, service delivery, policy analysis |
| Healthcare | Medium | DHBs, private providers | Clinical decision support, administration, patient communication |
| Professional services | Medium | Large firms | Knowledge management, document review, client insights |
| Agriculture/primary | Low-Medium | Large exporters | Supply chain, quality assessment, market intelligence |
| Education | Low-Medium | Universities | Student support, administration, research assistance |
Financial services and government are the clear leaders, driven by a combination of data maturity, regulatory pressure, and scale economics that justify platform investment.
Comparison: 2023 vs 2024 vs 2025
| Metric | 2023 | 2024 | 2025 |
|---|---|---|---|
| AI adoption (larger enterprises) | 48% | 61% | 72% |
| Formal governance frameworks | ~15% | ~34% | ~52% |
| Production AI capabilities (median, Operators) | 1 | 2 | 3-4 |
| Average first-capability investment | $80-150K | $100-200K | $120-250K |
| AI-specific board reporting | Rare | ~20% | ~40% |
| Foundation/platform approach | <5% | ~10% | ~15% |
NZ Enterprise AI Adoption Over Time (%)
Source: NZTech, AI Readiness in Aotearoa 2023-2025; RIVER Group
The trend is clear: steady progress across every dimension. But the distribution matters more than the average. The top 15% are compounding their advantage while the bottom 28% haven't started.
The Australia Comparison
NZ and Australia share similar enterprise challenges but differ in scale:
| Dimension | NZ | Australia |
|---|---|---|
| Adoption rate | 72% | 78% |
| Average investment | 0.6× AU equivalent | 1× (baseline) |
| Governance maturity | Comparable | Slightly ahead (regulated sectors) |
| Talent depth | Shallower pool | Deeper but more competitive |
| Government framework | Principles-based | More prescriptive |
| Speed of execution | Faster (smaller, fewer layers) | Slower (more complex approvals) |
NZ's structural advantages (smaller organisations, shorter decision chains, pragmatic leadership) continue to offset the scale disadvantage. NZ enterprises that commit to AI move faster than their Australian counterparts. The challenge is commitment, not capability.
Forward Look: 2026
Based on current trajectories and the engagements we're seeing:
Prediction 1: The Operator tier will reach 25%. Organisations that started building foundations in 2024-2025 will reach operational maturity in 2026. The middle tier will continue growing, but the top tier will grow faster.
Prediction 2: AI governance will become a procurement requirement. Government and large enterprise buyers will start requiring AI governance frameworks from their suppliers. This will drive governance adoption through supply chains, not just internal motivation.
Prediction 3: Agentic AI will move from concept to early production. Agentic AI (systems that take actions, not just answer questions) will see its first NZ enterprise deployments in 2026. Claims processing, compliance monitoring, and customer service will be the early use cases.
Prediction 4: The talent gap will shift again. The next bottleneck will be AI operations: teams that can monitor, maintain, and continuously improve production AI systems. Building AI is becoming easier; running AI reliably is the emerging challenge.
Prediction 5: Consolidation in the NZ AI services market. The 50+ firms offering AI consulting in NZ will consolidate. Enterprises will preference partners with proven delivery track records and compound platform capability over one-off consultants.
- Is NZ keeping pace with global AI adoption?
- In terms of awareness and intent, yes. In terms of operational deployment, NZ is slightly behind the US and UK but comparable to Australia and ahead of many European markets. NZ's advantage is execution speed. When we commit, we move fast. The challenge is that too many organisations are still in the commitment phase.
- What should NZ enterprises prioritise in 2026?
- If you're an Explorer: start a structured AI discovery immediately. If you're a Builder: invest in foundation infrastructure before your next capability. If you're an Operator: focus on organisational AI literacy and agentic AI readiness. Everyone: ensure your governance framework meets the NZ Government AI Framework principles.
- How does Māori data sovereignty apply to enterprise AI?
- Te Tiriti o Waitangi principles extend to AI systems that process data related to Māori communities, health, education, or cultural knowledge. Organisations should engage with iwi data governance principles, ensure appropriate controls on culturally sensitive data, and consider the implications of AI-generated outputs that affect Māori. This isn't a compliance checkbox. It's a responsibility.