New Zealand's AI ecosystem is small but growing. Twelve months after ChatGPT launched, the landscape has shifted from curiosity to early action. Here's a snapshot of who's building, who's buying, and where the gaps remain.
The Landscape
Enterprise Adoption
NZ enterprise AI adoption in late 2023 sits somewhere between "exploring" and "piloting." Most large organisations have at least one AI initiative underway. Very few have AI in production at scale.
The pattern varies by sector:
Financial services. The most active sector. Banks and insurers have data, they have compliance requirements that AI can help with, and they have the IT budgets to invest. Use cases: document processing, claims triage, customer service augmentation, regulatory compliance. Several NZ financial institutions have pilots running. A handful have early production deployments.
Government. High interest, cautious action. Government agencies are exploring AI for service delivery, policy analysis, and internal operations. The caution is justified - government data carries additional sensitivity, Te Tiriti obligations apply, and public sector procurement moves deliberately. The AI Centre of Excellence within GCDO is coordinating, but agency-level adoption is uneven.
Health. Emerging interest, significant constraints. The potential is enormous but the barriers are real: data sovereignty, clinical safety, regulatory requirements, workforce readiness. Most activity is in administrative applications rather than clinical AI.
Professional services. Law firms and consultancies are experimenting heavily with AI for document review, research, and content generation. Adoption is typically individual-driven rather than firm-wide.
42%
of NZ enterprises had at least one AI initiative underway by late 2023
Source: RIVER Group, NZ Enterprise AI Survey, November 2023
NZ Enterprise AI Adoption by Sector, Late 2023
Source: RIVER Group, NZ Enterprise AI Survey, November 2023
The Builder Ecosystem
NZ has a small but capable community of organisations building AI solutions:
Research institutions. University of Auckland, Victoria University of Wellington, and others have active AI research programmes. The challenge is translating research into enterprise application - a gap that exists globally but is more acute in a small market.
Startups. A growing number of NZ-based AI startups, primarily focused on specific vertical applications. The talent constraint is real - NZ's small pool of AI engineers means startups compete with each other and with international opportunities.
Consultancies and integrators. Several NZ technology consultancies (ourselves included) are building AI advisory and delivery capability. The demand is outpacing supply.
Global vendors with NZ presence. Microsoft, Google, AWS, and others have NZ teams, but the AI capability tends to be centralised globally. NZ customers often deal with local relationship managers backed by offshore technical resources.
Where the Gaps Are
Sovereign AI infrastructure. NZ has no significant AI compute infrastructure. All major AI processing happens offshore, primarily in the US. For sensitive use cases (health, government, indigenous data), this is a material constraint.
AI talent. The supply of AI engineers, data scientists, and ML specialists in NZ is far below demand. This is a global problem but acute in a small market where the talent pool is measured in hundreds, not thousands.
Industry-specific solutions. Most available AI solutions are built for global markets and don't account for NZ-specific requirements: te reo Māori, NZ regulatory frameworks, NZ industry terminology, small-market economics.
Governance capability. Very few NZ organisations have the internal capability to design and implement AI governance frameworks. This is a consulting opportunity and an ecosystem gap.
Māori AI capability. Indigenous AI - systems designed with and for Māori communities, respecting data sovereignty and cultural governance - is almost entirely absent. This represents both a responsibility and an opportunity for the NZ ecosystem.
NZ AI Ecosystem Gaps: Severity Assessment
Source: RIVER Group, NZ AI Ecosystem Analysis, November 2023
What We're Observing
The "Wait and See" Problem
Too many NZ enterprises are in "wait and see" mode. Watching international developments, reading Gartner reports, attending conferences. Not building, not piloting, not developing internal capability.
The problem: AI capability compounds. Organisations that start experimenting now - even imperfectly - are building data pipelines, training teams, and developing governance frameworks that will accelerate everything they do next year. Organisations that wait will face a larger gap to close.
The Vendor Dependency Risk
NZ enterprises are overwhelmingly dependent on international vendors for AI capability. This is fine for commodity AI applications. But for applications that require NZ-specific knowledge, NZ-hosted processing, or NZ cultural competency, vendor dependency is a strategic risk.
The NZ ecosystem needs more organisations building AI solutions for NZ contexts. Not reinventing GPT-4 - building the enterprise wrappers, the data pipelines, the governance frameworks, and the domain-specific applications that make global AI models useful in a New Zealand context.
The Collaboration Opportunity
NZ is small enough that meaningful ecosystem coordination is possible. Government, enterprise, research, and the Māori data sovereignty community could align on shared infrastructure, shared standards, and shared capability development in a way that larger markets can't.
This isn't happening yet. But the ingredients are present.
What We'd Recommend
For the ecosystem broadly:
- Invest in sovereign AI infrastructure. This is a public-private opportunity.
- Develop AI talent pathways. Not just technical skills - governance, ethics, and domain expertise applied to AI.
- Build NZ-specific solutions. Don't wait for global vendors to solve NZ problems.
- Centre indigenous perspectives. Māori data sovereignty isn't a constraint on innovation. It's a design principle that produces better AI.
- Coordinate. Small markets benefit from collaboration more than from competition.