Skip to content

Māori Data and AI – guidance for business

About this guidance

This guidance is intended to be high-level, introducing basic concepts to stimulate further investigation and thinking. There are various pieces of work underway to develop more detailed and practical guidance, which will replace this guidance when published.

Māori Data and AI

Artificial Intelligence (AI) is built on data. Therefore, it is important to understand Māori data before beginning to develop, train, or use AI that incorporates Māori data.

What is Māori Data

Digital or digitisable information or knowledge that is about or from Māori people, language, culture, resources, or environments. For many Māori, data is considered a Taonga and subject to Māori Governance.

Data as a Taonga

Taonga are valued and treasured things, both tangible and intangible.

This means extra care and consideration needs to be taken if you are producing, using, or handling Māori data in your business.

What to do practically

Do no harm and ensure Māori benefit from the use of their data

Carefully consider how any data collection and use will benefit Māori both now and in the future. When using AI with Māori data, the risks are more than just those concerned with data ownership and control – often referred to as Māori data sovereignty. There is a higher likelihood that bias, discrimination, inaccuracy, and other harms could materialise for Māori in your AI use. When you use data for AI that is collected as part of your business-as-usual administrative or operational processes, be aware that risk may be increased due to the potential bias within this type of data. This will likely introduce risks from the outset.

Here are some practical tips:

  • Ensure diverse teams, expertise and skill sets are involved in the development of AI. This will help you identify any potential issues.
  • Identify potential risks or harms that could occur, such as discrimination or stigmatisation, such as from biased or unrepresentative data.
  • Have a plan to mitigate these risks.
  • Have an exit strategy to transition away from the tool if it proves problematic.
  • Always have a person in the process to assess for unintended consequences.
  • Consider that risks will be difficult to fully identify without Māori expertise and perspective as part of your analysis and could lead to paternalistic decisions that fail to identify risks for Māori.
Identify and understand what Māori data you collect or use

Respect that Māori may view their data differently from you. Seek to understand how Māori consider their information and what they view as sensitive or tapu (sacred or restricted). Anything that is tapu will require more extensive measures to appropriately manage that data.

Here are some practical tips:

  • Identify Māori data and work with Māori to describe it in culturally relevant ways.
  • Understand why you need to use Māori data to begin with. This will help inform how best to work with it.
  • Is there data missing that Māori consider important and necessary? Work with Māori to collect and incorporate this data and maintain important contextual information.
  • Understand whether the data is representative and of appropriate quality. If it is not, understand the limitations this presents for the intended use and analysis and report those limitations to inform interpretation.
Be transparent and explain how AI uses Māori data

Be able to explain how Māori data is used, for what purpose, and how it was used by the AI. Document all decision-making and data usage in ways that are transparent and accessible.

What is Māori Data Sovereignty

Māori Data Sovereignty is the right for Māori to collect, own, and choose who uses their data.  Māori Data Governance incorporates principles, structures, accountability mechanisms, legal instruments, and policies through which Māori exercise control over their data.

What to do practically

Invite representation of Māori views in decision-making

If your use or development of AI systems uses Māori data or impacts Māori, then it is appropriate to invite representation of Māori views in decision-making.

  • If you don’t have any Māori partnerships, you should look for ways to create them.
  • Involve Māori early in decisions and the development of AI, so they can guide and decide what happens with their data.
  • Check with Māori that the AI is interpreting Māori data within the correct context and is culturally appropriate.
Set clear boundaries on Māori data use

Understand what data Māori may be willing to share and under what conditions, and for what purposes.

  • Establish clear boundaries for what types of Māori data can be used for different use cases and seek informed consent before beginning.
  • Many Māori view data as belonging not just to them but to their whānau, hapū and iwi. Consider ways to protect this collective information.
  • Avoid using Māori data for AI training without explicit permissions and ensure transparency in how data is processed.
Grow your business capability

Seek support and training to help your business govern, manage, use and interpret Māori data in culturally safe ways. Some starting resources are included at the end of this guidance.

Guiding values

A lot of Māori businesses, big tech companies, and Māori data experts have developed guiding values or tikanga to help support people working with Māori data. While there is not one list, the following values are commonly used:

Mana whakahaere – Authority

Mana whakahaere speaks to the ability for Māori to have sovereignty, leadership, and autonomy to make decisions and determine the direction you take. Make sure you have the right subject matter expertise involved. Appropriate governance or advisory groups can help to ensure you have the right authority and are using the data in a way that is appropriate.

Kaitiakitanga – Guardianship

This is the practice of applying careful, responsible, and ethical practices when using Māori data. This value talks to the care, protection, and understanding of the potential impact on wider communities when you are using data about Māori. Often, this aligns with the concepts of noa and tapu, and of balancing benefits and risks. As part of your guardianship role, understand the risk and sensitivity of Māori data. Regardless of how you are using the data, there is an obligation to keep the data safe, protect its privacy and confidentiality, and ensure appropriate use. 

Manaakitanga - Respect

This value goes alongside kaitiakitanga and is about creating an environment where Māori are genuinely welcomed and listened to. Take a collaborative approach to develop solutions alongside your partners.  

Whakapapa – Relationships

Understand where the data comes from and the relationships with the communities who will be most present in that data.  Accurate metadata (information about the data) should, at minimum, provide information about where the data comes from, the purpose(s) for its collection, the context of its collection, and the parties involved.

This also relates to the idea of mauri and understanding how the data will be used and the extent to which it will transform from its original collection purpose.

He waka eke noa – Collective benefit

When using Māori data, you should consider the benefits to communities the data comes from. The data should be used to open and unlock new opportunities for these communities. From an AI and data perspective, you can apply this principle to understanding insights and make informed decisions that benefit the people who share their data with you.

The misuse or exploitation of Māori data can be avoided with appropriate safeguards and cultural considerations. 

Start the journey

Many businesses in Aotearoa are small and may not have the right capability, capacity, and knowledge. Others may have the people and resources but not the knowledge and experience. The best approach is simply to start somewhere. Seek support to learn, gain skills, deepen understanding, and be able to assess risks. Take the steps that you can now, and work towards understanding and implementing further steps when you are ready. The most important thing to do is just start.

Here are some resources that might be useful:

Top