What is AI governance?
AI governance is the system of decisions, accountabilities, policies and controls an organisation uses to make sure its use of artificial intelligence is lawful, ethical, safe and trusted. In Aotearoa New Zealand, good AI governance must also give effect to Te Tiriti o Waitangi and to Māori authority over Māori data, knowledge and taonga.Our approach to AI governance
Most AI governance frameworks used in New Zealand are imported. They manage technical and commercial risk well, but treat Māori rights as a consultation step or a footnote. Our approach starts from five principles.1. Te Tiriti o Waitangi is a threshold, not a checkbox
Te Tiriti obligations are tested before an AI system is approved, not added afterwards as an engagement exercise.2. Architecture, not decoration
We test whether the Māori provisions in an AI policy actually change decisions, controls and accountability, or whether they only appear in the introduction. If a Māori principle can be removed without changing how the system operates, it was decoration.3. Sovereignty, not just residency
Storing data in New Zealand is data residency. Māori Data Sovereignty is about who controls access, use, reuse, model training and deletion. We assess both, drawing on the Waitangi Tribunal’s findings in WAI 2522.4. Rangatiratanga sits with iwi and hapū
Engagement with a single pan-Māori body does not discharge obligations to the iwi and hapū whose people, data and taonga are affected.5. People stay accountable
Every automated or AI-assisted decision should have a named human who is accountable for it, with clear rules about when people are in the loop, on the loop, or answerable for the outcome.AI governance services
-
- AI governance frameworks and policy: AI policies, acceptable-use rules, risk appetite statements and decision gates that work in practice.
-
- Board and executive advisory: briefings, director education and independent advice on AI strategy, risk and accountability.
-
- AI risk and impact assessments: assessments of cultural, ethical, privacy, legal and social risk before deployment, including algorithmic bias and impacts on Māori.
-
- Te Tiriti and Māori Data Sovereignty reviews: independent review of AI programmes, procurement and vendor contracts against Te Tiriti obligations and Māori Data Sovereignty.
-
- Generative AI and mātauranga Māori: advice on training data, scraping, cultural intellectual property and the appropriate use of te reo Māori and mātauranga Māori in AI systems.
-
- Automated decision-making and human oversight: oversight models, escalation paths and accountability for automated decisions in public services.
-
- Policy critique and submissions: independent analysis of government AI guidance, legislation and standards.
Who we work with
Boards, government agencies, iwi and hapū, Māori organisations, universities, research institutes and technology companies, both those deploying AI and those building it. We are based in Ōtautahi Christchurch and work nationwide.Experience
Dr Karaitiana Taiuru has worked on Māori technology sovereignty since 2006. He co-founded the Institute of AI Governance, is a Member of the Institute of Directors, and holds Standards New Zealand appointments to international ISO/IEC committees, including SC 42 (Artificial Intelligence). This gives our advice an uncommon mix of international standards practice, governance experience and Māori rights expertise. View professional credentials →Original frameworks
Our AI governance work is underpinned by original Māori AI frameworks developed by Dr Taiuru, including He Tangata, He Karetao, He Ātārangi, a kaupapa Māori framework for artificial intelligence.AI governance FAQ
Why does AI governance in New Zealand need to consider Te Tiriti o Waitangi?
AI systems use data about Māori people, communities and knowledge, and their decisions affect Māori outcomes. Te Tiriti o Waitangi guarantees Māori authority over their taonga, which the Waitangi Tribunal’s WAI 2522 findings recognise as including data. AI governance that ignores Te Tiriti exposes organisations to legal, reputational and social licence risk.What is the difference between data residency and Māori Data Sovereignty?
Data residency is about where data is physically stored. Māori Data Sovereignty is about who has authority over that data: who can access it, use it, reuse it, train AI models on it and delete it. Data can be stored in New Zealand and still be outside Māori control.What is the architecture-versus-decoration test?
It is a test of whether Māori provisions in an AI policy or system are structurally operative. If removing a Māori principle would not change any decision, control or accountability, the principle is decoration rather than architecture.Is consulting one national Māori body enough?
No. Rangatiratanga is held by iwi and hapū. Organisations should engage with the iwi and hapū whose people, data and taonga are affected by an AI system, rather than relying on a single pan-Māori body.Who should be accountable for AI decisions?
A named person. Every automated or AI-assisted decision should have a human who is accountable for its outcome, with clear rules about when people review decisions before they take effect, monitor them, or answer for them afterwards.Does Taiuru & Associates sell AI software?
No. We provide independent advice. We do not sell software, which means our recommendations are not tied to any vendor or platform.Recent AI governance analysis
All AI governance articles →Talk to us
If your organisation is deploying AI, writing an AI policy, or wants an independent view on whether its existing approach would stand up to scrutiny, we can help. Contact Taiuru & Associates →Last updated: 8 October 2026