Table of Contents

• 1. Overview
• 2. Introduction
• 2.1 Aotearoa New Zealand research context
• 2.2 What is generative artificial intelligence?
• 2.3 GenAI and mātauranga Māori
• 3. The Guidelines
• 3.1 Principles
• 3.2 Guidelines for the best-practice use of GenAI in research in Aotearoa New Zealand
• 4. Do you have feedback on these guidelines?
• 5. Other frameworks and resources
• 6. References
• Appendix 1: Definitions
• Appendix 2: GenAI-relevant legal frameworks and guidelines

1. Overview

1.1 Royal Society Te Apārangi

Royal Society Te Apārangi is an independent, not-for-profit organisation that promotes and serves the Aotearoa New Zealand research sector.

One of the Society’s parliamentary remits is to provide infrastructure and other support for the professional needs and development of science, technology, and humanities researchers. This includes research practice guidance.

1.2 Advisory group

These guidelines were developed in collaboration with:

• Professor Ali Knott, Te Herenga Waka – Victoria University of Wellington
• Associate Professor Te Taka Keegan, The University of Waikato, Te Whare Wānanga o Waikato
• Professor Giselle Byrnes, Massey University – Te Kunenga ki Pūrehuroa
• Dr Simon McCallum, Te Herenga Waka – Victoria University of Wellington
• Associate Professor Mahsa Mohaghegh, Auckland University of Technology, Te Wānanga Aronui o Tāmaki Makau Rau

Our appreciation to Associate Professor Mele Taumoepeau (THW–VUW) and Lynell Tuffery Huria of Kāhui Legal. Thank you very much to the group of 24 researchers who attended a workshop to discuss and review the guidelines.

1.3 Purpose of these guidelines

Our aim is for researchers in Aotearoa New Zealand to draw on these guidelines to help navigate a rapidly evolving landscape, enabling them to leverage the innovation and evident benefits that generative artificial intelligence (GenAI) offers, while at the same time being aware of risks and the need to ensure GenAI is used ethically with humans in control.

These are suggested guidelines, and it is expected that individuals and organisations will tailor these guidelines for their own purposes and contexts.

1.4 Who are these guidelines for?

These guidelines are for all researchers in Aotearoa New Zealand who use, or are thinking of using, GenAI tools. This includes those in universities, wānanga, Crown Research Institutes, Māori research organisations, and Pacific research organisations, along with industry-based researchers and independent researchers.

1.5 Scope

The focus of this iteration of the guidelines is on GenAI agents that are used to create new text, images, audio, or video.    

Out of scope are supervised models, unsupervised models, and reinforcement learning (1). The use of GenAI tools and platforms in education (teaching and learning) and the formal use of GenAI by research institutions and research funders are also out of scope, in addition to the wider societal applications of this technology.

2. Introduction

Generative artificial intelligence (GenAI) is one of the technological tools that researchers are increasingly utilising. The use of GenAI provides many opportunities to advance and accelerate discovery and the pace of research, fostering new ideas and unleashing innovation. GenAI tools can save time, enhance efficiency and creativity, improve accessibility (see (2)), and simplify data analysis and management (3,4). However, these tools also harbour ethical risks to research integrity [1]. In Aotearoa New Zealand, these include risks to fulfilling obligations to te Tiriti o Waitangi and Māori data sovereignty. The upholding of Pacific data sovereignty, too, faces threats.

These guidelines suggest principles and recommendations to support the best-practice use of GenAI in research. They are framed by recognition of te Tiriti o Waitangi, fairness, equity of access, and consideration of diverse perspectives and users. The guidelines aim to provide innovative and possibly distinctive understandings of GenAI application in a research context where culturally diverse knowledge systems interweave.

The pace of change of GenAI technology means that these guidelines must be understood as dynamic and accurate as at the time of publication (June 2025).

[1] GenAI use in research has multiple limitations, risks, challenges, and sensitivities with regards to information. These include accuracy, disinformation, deception, data collection, ownership, authorship, attribution, accountability, transparency, ethics, fairness, bias, discrimination, user control, confidentiality, privacy, and security (5,6).

2.1 Aotearoa New Zealand research context

These guidelines are specific to Aotearoa New Zealand’s distinct research landscape, which is underpinned by our unique cultural context and history – in addition to being part of an international research community.

Research in Aotearoa New Zealand is informed by obligations to te Tiriti o Waitangi (7) and to the tikanga of te ao Māori, the Māori worldview. Protection of taonga Māori, which includes te reo Māori, mātauranga Māori, and Māori data, is of key importance (8). Kaupapa Māori rangahau is tikanga led and community focused (9). Pacific research approaches and priorities are similar, yet distinct (10).

Aotearoa New Zealand’s te Tiriti o Waitangi foundations, along with our physical location and cultural connection with Te Moana-nui-a-Kiwa, the Pacific region, give rise to research that not only addresses a range of problems and creates new knowledge, but is typically people-centric and measured by the ways it benefits and uplifts communities.

2.2 What is generative artificial intelligence?

Artificial intelligence (AI) refers to machines programmed to mimic human intelligence. AI is used to analyse data to recognise patterns and make predictions based on those patterns. 

Generative artificial intelligence (GenAI) is a subset of AI. In response to human instructions or ‘prompts’, GenAI can create new content, such as text, images, video, audio, and code. It does this through being trained on very large collections of existing content – scraped from the internet and other sources. GenAI technologies are now revolutionary in the high quality of their content: it can be difficult to distinguish from human-generated content.

2.3 GenAI and mātauranga Māori

Te Tiriti o Waitangi is a partnership between te iwi Māori and the Crown, and an acknowledgement of tino rangatiratanga and obligations to taonga Māori.

GenAI tools are of concern because they can draw on mātauranga Māori – a taonga that includes te reo Māori and Māori data – without Māori knowledge or consent, circumventing te Tiriti o Waitangi obligations and ignoring te Tiriti o Waitangi-compliant research principles, such as tika and mana.

For example, although most existing GenAI tools that generate te reo Māori can give responses as if from a te ao Māori perspective, the majority are lacking in cultural and linguistic nuance, dialect, and iwi-specific perspectives. Such tools are enabled by the disregarding of principles of Māori data sovereignty (11). Researchers who use GenAI to generate te reo Māori or Māori data must therefore consider their obligations and the limitations of GenAI tools.

3. The Guidelines

These guidelines give recommendations for the best-practice use of GenAI in research in Aotearoa New Zealand. They draw on the European Commission’s Living Guidelines on the Responsible Use of Generative AI in Research (3)[2] and UNESCO’s Guidance for Generative AI in Education and Research (12) [3], and are informed by current thinking at Aotearoa New Zealand universities (13) regarding how to use GenAI responsibly and ethically.

[2] CC-BY 4.0 (1)
[3] CC-BY-SA 3.0 IGO

3.1 Principles

3.1.1 Respect the interests and perspectives of Māori as articulated in te Tiriti o Waitangi (7)

Article Two of te Tiriti o Waitangi states that Māori have 'tino rangatiratanga' over their 'taonga katoa' (7). This article infers that Māori should have control over taonga Māori, including Māori knowledge, Māori language, and Māori data.

Te Tiriti o Waitangi and its principles require consideration on an ongoing basis as the use of GenAI evolves (14).

3.1.2 Follow key principles of research integrity

As articulated in the Research Charter for Aotearoa New Zealand (15) and the Code of Professional Standards and Ethics in Science, Technology, and the Humanities (16)[4]. For Pacific research guidelines, see (17).

[4] The Māori values were developed in 2019 by a Royal Society Te Apārangi working group comprising Professor Tahu Kukutai, Associate Professor Maui Hudson, Associate Professor Maria Bargh, Associate Professor Hēmi Whaanga, Dr Hinemoa Elder, Dr Dan Hikuroa, Marino Tahi, and Moe Milne for inclusion in the Society’s Code of Professional Standards and Ethics and the Research Charter.

3.2 Guidelines for the best-practice use of GenAI in research in Aotearoa New Zealand

To responsibly and ethically use GenAI in research, researchers should:

3.2.1 Acknowledge and respect that Māori data, including mātauranga Māori and te reo Māori, are taonga that Māori have te Tiriti o Waitangi-afforded rights to govern and protect.

• Researchers’ decisions about the use of GenAI adhere to principles of Māori data sovereignty (11).       
• Researchers using GenAI tools that incorporate Māori data are responsible for ensuring that no GenAI tool will assume any proprietary interest of Māori data, including mātauranga Māori and te reo Māori (13).    
• Researchers with intentions to commercialise any outcome from the use of a GenAI tool that incorporates Māori data require free prior informed consent from Māori kaitiaki on access and use of the Māori data, and fair and equitable benefit-sharing terms (13).

3.2.2 Be responsible for research outputs

• Researchers are accountable for the integrity of the content [5] generated by or with the support of GenAI tools.
• Researchers are expected to maintain a critical approach to using the output produced by generative AI and are responsive to the tools’ limitations, such as cultural limitations, bias, hallucinations [6], and inaccuracies.
• GenAI systems are neither authors nor co-authors. Authorship implies agency and responsibility, and therefore lies with human researchers.

[5] Content should not be fabricated, falsified or plagiarised. More details in Section 3.1 of The European Code of Conduct for Research Integrity, http://www.doi.org/10.26356/ECOC, ALLEA 2023.

[6] ‘Hallucination’ refers to the generation of false, nonsensical, or inaccurate information by large language models (LLMs) or other GenAI systems.

3.2.3 Use GenAI transparently

• Researchers indicate which GenAI tools have been used substantially [7] in their research processes. Acknowledgement of the tool could include, for instance, the name, version, date, and how it was used and affected the research process. If relevant, researchers make the input (prompts) and output available, in line with open research principles.
• Researchers take into account the stochastic (random) nature of GenAI tools, which is the tendency to produce different outputs from the same input.
• Researchers disclose or discuss the limitations of GenAI tools used, including possible biases in the generated content, as well as possible mitigation measures.
• Transparency requirements are becoming more formal in academic publishing, as discussed in Section 3.2.9.

[7] Examples of uses that could have a substantial impact include interpreting data analysis, writing a literature review, identifying research gaps, formulating research aims, and developing hypotheses.

3.2.4 Pay particular attention to issues related to:

• data sovereignty (8), which includes Māori data sovereignty (8), Pacific data sovereignty (17), and Indigenous data sovereignty (18)
• privacy
• confidentiality
• intellectual property rights
  
  when sharing sensitive or protected information with GenAI tools.

• Researchers understand implications related to Māori data sovereignty (8), Pacific data sovereignty (17), and Indigenous data sovereignty (18).
• Researchers understand the technical and ethical implications regarding privacy, confidentiality, and intellectual property rights. Researchers will check, for example, the privacy options of the tools, who is managing the tool (public or private institutions, companies, etc), where the tool is running, and implications for any information uploaded. This could range from closed environments, hosting on a third-party infrastructure with guaranteed privacy, to open internet-accessible platforms (see Appendix 1). Researchers will ensure all confidentiality obligations are upheld and intellectual property rights are acknowledged, respected, and protected.
• Researchers remain mindful that generated or uploaded input (text, data, prompts, images, etc) could be used for other purposes, such as the training of GenAI models. They protect unpublished or sensitive work by taking care not to upload it into an online AI system unless there are assurances that the data will not be stored and/or reused.
• Researchers respect the providers of the raw data, which includes ensuring their rights are not infringed and that they benefit from the use of their data. Researchers take care not to provide personal data to online GenAI systems unless the data subject (individual or collective) has given their free prior informed consent and researchers have a clear goal for which the personal data are to be used, so compliance with data protection rules is ensured [8].

[8] https://commission.europa.eu/law/law-topic/data-protection/eu-data-protection-rules_en;  https://www.privacy.org.nz/privacy-act-2020/privacy-principles/

 
3.2.5 Respect applicable New Zealand legislation and Waitangi Tribunal recommendations, as well as international legislation, as in regular research activities (see Appendix 2)

• The output produced by GenAI should be sensitive to the legal protection of intellectual property rights and personal data.
• The Waitangi Tribunal reports WAI 2522 and WAI 262 have recommendations that refer to obligations associated with the protection of mātauranga Māori, including Māori data and te reo Māori (19,20).
• When using outputs from GenAI, researchers pay attention to the potential for plagiarism (text, code, images, etc). Researchers validate the sources of information and respect others’ authorship, citing work where appropriate.
• Consideration should be given to international legal frameworks and principles, such as the United Nations Declaration on the Rights of Indigenous Peoples (21) and obligations to the protection of Indigenous knowledge, including Indigenous data and Indigenous languages (see Article 31 (21)).

3.2.6 Maintain a critical technical and ethical approach to using GenAI

• Researchers are aware of GenAI limitations, such as hallucinations and bias.
• Researchers take steps to mitigate unintended negative consequences of GenAI use. They also take steps to utilise GenAI to improve equity, inclusion, and accessibility – benefiting communities with diverse abilities and needs.
• Researchers are mindful of environmental impacts of GenAI:
  • Use of GenAI is energy intensive and raises considerations regarding sustainability.
  • Honouring te Tiriti o Waitangi requires protection of te taiao, the environment.

3.2.7 Continually learn how to maximise the benefits of GenAI tools

• Researchers should stay up to date with best practices, including by undertaking training. They are also encouraged to share their skills with colleagues and other stakeholders.
• GenAI literacy and access to current tools need to be equitably available to everyone, and researchers have a responsibility to take steps to ensure access to others.

3.2.8 Be particularly deliberate and transparent if using AI and GenAI tools substantially [9] in sensitive activities that could impact other researchers or organisations, such as peer reviews and evaluations of research proposals

• Use of GenAI tools creates the potential for unfair assessment as a result of these tools’ limitations (such as hallucinations and bias).
• Avoiding the use of GenAI safeguards the original unpublished work of fellow researchers from potential exposure or inclusion in a GenAI model.

3.2.9 Be aware of publishers’ policies around Gen AI

• Policies on GenAI from publishers and other academic venues differ and are changing frequently.
• Researchers should also refer to the policies of the organisations by whom they are employed and with which they are interacting.

[9] For example, using GenAI to search background information for a review is not a substantial use, while delegating the assessment of a paper is a substantial use.

4. Do you have feedback on these guidelines?

GenAI may create possibilities and risks that can be hardly anticipated and fully comprehended today. Given the dynamic, fast-evolving nature of GenAI technology, ongoing feedback from the research community is encouraged to keep these guidelines up to date.

Practical guidance for applying these principles in practice throughout the research cycle is in development.

Email: expertadvice@royalsociety.org.nz

5. Other frameworks and resources

• European Commission: ‘Living guidelines on the Responsible Use of Generative AI in Research’, March 2024: https://research-and-innovation.ec.europa.eu/document/2b6cf7e5-36ac-41cb-aab5-0d32050143dc_en
• Guidance for generative AI in education and research | UNESCO: https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research
• Royal Society Te Apārangi. Mana Raraunga Data Sovereignty. Dec 2023: https://www.royalsociety.org.nz/what-we-do/our-expert-advice/all-expert-advice-papers/mana-raraunga-data-sovereignty
• Royal Society Te Apārangi. Research Charter for Aotearoa New Zealand: https://www.royalsociety.org.nz/what-we-do/research-practice/research-charter/research-charter-aotearoa-new-zealand/#4955
• Royal Society Te Apārangi. Code of Professional Standards and Ethics in Science, Technology, and the Humanities: https://www.royalsociety.org.nz/who-we-are/our-rules-and-codes/code-of-professional-standards-and-ethics/code-of-professional-standards-and-ethics-in-science-technology-and-the-humanities/
• Prime Minister’s Chief Science Advisor, Dec 2023: Capturing the benefits of AI in Healthcare: https://www.pmcsa.ac.nz/artificial-intelligence-2/ai-in-healthcare
• Massey University Te Kuenenga ki Pūrehuroa, Use of Artificial Intelligence in Assessment Policy: https://www.massey.ac.nz/documents/1445/Use_of_Artificial_Intelligence_in_Assessment_Policy.pdf
• Massey University Te Kuenenga ki Pūrehuroa, Generative Artificial Intelligence (GenAI) Usage Guidelines for Staff: https://www.massey.ac.nz/documents/2046/Generative_Artificial_Intelligence_GenAI_Usage_Guidelines_for_Staff.pdf
• AI Forum NZ: Trustworthy AI in Aotearoa report: https://aiforum.org.nz/reports/trustworthy-ai-in-aotearoa-the-ai-principles
• EU: Trustworthy AI: Ethical principles for AI systems. From Ethics Guidelines for Trustworthy AI by the EU high-level expert group on AI: https://op.europa.eu/en/publication-detail/-/publication/d3988569-0434-11ea-8c1f-01aa75ed71a1
• AI Researchers Association: White paper: Aotearoa New Zealand Artificial Intelligence, A strategic approach 2024: https://www.airesearchers.nz/site_files/28243/upload_files/AotearoaNewZealandAI_AStrategicApproach_2024Edition_HIGHRES.pdf?dl=1
• New Zealand Public Service AI Framework
  https://www.digital.govt.nz/standards-and-guidance/technology-and-architecture/artificial-intelligence/public-service-artificial-intelligence-framework
• KU Leuven University: Using generative artificial intelligence as a researcher: https://research.kuleuven.be/en/integrity-ethics/integrity/practices/genAI
• Association for Computing Machinery (ACM): Principles for the development, deployment, and use of generative AI technologies: https://www.acm.org/binaries/content/assets/public-policy/ustpc-approved-generative-ai-principles
• International Association of Scientific, Technical and Medical Publishers (STM): Generative AI in Scholarly Communications – Ethical and practical guidelines for the use of Generative AI in the Publication Process https://www.stm-assoc.org/wp-content/uploads/STM-GENERATIVE-AI-PAPER-2023.pdf
  
  

6. References

1. Bowen JA, Watson CE. Teaching with AI: A Practical Guide to a New Era of Human Learning. American Association of Colleges and Universities. American Association of Colleges and Universities; 2024.

2. Morris MR. AI and accessibility. Communications of the ACM. 2020;63(6): 35–37. https://doi.org/10.1145/3356727.

3. European Commission. Living guidelines on the responsible use of generative AI in research | Research and innovation. In: 2024. https://research-and-innovation.ec.europa.eu/document/2b6cf7e5-36ac-41cb-aab5-0d32050143dc_en [Accessed 26th August 2024].

4. Byrnes G, Lovreglio R. Opinion: Navigating through quicksand? Using GenAI ethically. Te Kunenga ki Pūrehuroa - Massey University of New Zealand. https://www.massey.ac.nz/about/news/opinion-navigating-through-quicksand-using-genai-ethically/ [Accessed 26th August 2024].

5. Tabassi E. Artificial Intelligence Risk Management Framework (AI RMF 1.0). Report number: NIST AI 100-1, 2023 Jan [Accessed 11th February 2025]. p. NIST AI 100-1. https://doi.org/10.6028/NIST.AI.100-1. [Accessed 11th February 2025].

6. Palmer J. How to harness AI’s potential in research — responsibly and ethically. Nature. 2024;632(8027): 1181–1183. https://doi.org/10.1038/d41586-024-02762-2.

7. Te Tiriti o Waitangi, The Treaty of Waitangi. Archives New Zealand. https://archives.govt.nz/discover-our-stories/the-treaty-of-waitangi [Accessed 13th May 2020].

8. Royal Society Te Apārangi. Mana Raraunga Data Sovereignty. 2023 [Accessed 13th February 2025]. https://www.royalsociety.org.nz/assets/Mana-Raraunga-Data-Sovereignty-web-V1.pdf [Accessed 13th February 2025].

9. Smith LT. Kaupapa Māori research - Some Kaupapa Māori principles. In: Kaupapa Rangahau A Reader: A Collection of Readings from the Kaupapa Māori Research Workshop Series. Te Kotahi Research Institute; 2015. p. 46–52. https://researchcommons.waikato.ac.nz/entities/publication/571b43c8-5422-47ee-873c-40821c81f0bb [Accessed 31st January 2025].

10. Naepi S. Pacific Research Methodologies. Oxford Research Encyclopedia of Education. https://doi.org/10.1093/acrefore/9780190264093.013.566

11. Te Mana Raraunga. Te Mana Raraunga Māori Data Sovereignty Principles. 2018 [Accessed 16th May 2023]. https://static1.squarespace.com/static/58e9b10f9de4bb8d1fb5ebbc/t/5bda208b4ae237cd89ee16e9/1541021836126/TMR+Ma%CC%84ori+Data+Sovereignty+Principles+Oct+2018.pdf [Accessed 16th May 2023].

12. Guidance for generative AI in education and research | UNESCO. https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research [Accessed 13th February 2025].

13. Te Kunenga ki Pūrehuroa Massey University. Generative Artificial Intelligence (GenAI) Usage Guidelines for Staff. 2024. https://www.massey.ac.nz/documents/2046/Generative_Artificial_Intelligence_GenAI_Usage_Guidelines_for_Staff.pdf

14. Prime Minister’s Chief Science Advisor, Kaitohutohu Mātanga Pūtaiao Matua ki te Pirimia. Capturing the benefits of AI in healthcare for Aotearoa New Zealand. 2023 Oct [Accessed 17th October 2024]. https://bpb-ap-se2.wpmucdn.com/blogs.auckland.ac.nz/dist/f/688/files/2023/12/PMCSA-AI-healthcare-LONG-REPORT-FINAL-pdf-version-v3.pdf [Accessed 17th October 2024].

15. Research Charter for Aotearoa New Zealand. Royal Society Te Apārangi. https://www.royalsociety.org.nz/what-we-do/research-practice/research-charter/research-charter-aotearoa-new-zealand/ [Accessed 2nd September 2024].

16. Royal Society Te Apārangi. Code of Professional Standards and Ethics in Science, Technology, and the Humanities. https://www.royalsociety.org.nz/who-we-are/our-rules-and-codes/code-of-professional-standards-and-ethics/code-of-professional-standards-and-ethics-in-science-technology-and-the-humanities/ [Accessed 18th October 2024].

17. Pacific Data Sovereignty: Reclaiming Pacific Data. https://pacificdatasovereignty.com/ [Accessed 13th February 2025].

18. GIDA. Indigenous Peoples’ Rights in Data. Global Indigenous Data Alliance. https://www.gida-global.org/data-rights [Accessed 13th February 2025].

19. Waitangi Tribunal. Wai 262 Report: Protecting and Nurturing Taonga Māori. https://www.wai262.nz/ko-aotearoa-tenei [Accessed 26th October 2023].

20. Waitangi Tribunal. Wai 2522. https://www.justice.govt.nz/assets/WT-Wai-2522-2.5.0009-TPPA.pdf [Accessed 31st August 2023].

21. United Nations. United Nations Declaration on the Rights of Indigenous Peoples. United Nations; 2008. https://www.un.org/esa/socdev/unpfii/documents/DRIPS_en.pdf [Accessed 2nd February 2023].

Appendix 1: Definitions

Research and descriptions of the terms in Appendix 1 have been supervised by Dr Simon McCallum using a variety of AI tools including ChatGPT o3, Gemini 1.5 pro with Deep Research, and Perplexity.ai.  All content was reviewed and corrected to ensure accuracy.

Closed environments

A closed environment is a computing or communication ecosystem where access and interaction are tightly controlled and limited. In such an environment, the system’s components (programmes), data (email, files, web access), and operations (processes and meetings) are limited to a defined boundary with little or no interaction with external systems.

Key characteristics:

• Restricted access: Only approved users, applications, or devices can access the environment.
• Centralised control: A single entity or a tightly governed group manages updates, security, and system modifications.
• Isolation: The environment is typically isolated from the broader internet or external networks, reducing exposure to external threats.
• Enhanced security: The limited interface with external systems can reduce vulnerabilities and simplify compliance with strict security or regulatory requirements.

Example contexts:

• Enterprise networks with strict firewall rules.
• Specialised industrial control systems that operate in isolated networks.

Hosting on a third-party infrastructure with guaranteed privacy

This refers to a situation where an organisation places its data or applications on servers or cloud services managed by an external provider, with the assurance that strong measures are in place to protect the confidentiality and integrity of the data. Despite not controlling the physical infrastructure, the organisation receives contractual or technical guarantees that its information will remain private. This is the common approach by most companies and anyone using Cloud services such as Office365, Amazon AWS, or Google Drive.

 Key characteristics:

• Third-party management: The physical or virtual infrastructure is owned and maintained by an external service provider.
• Privacy guarantees: The service provider commits (often through contractual agreements and technical safeguards) to ensure data confidentiality. This might involve end-to-end encryption, strict access controls, and compliance with relevant privacy regulations.
• Risk management: While relying on a third party, the organisation mitigates risks through service-level agreements (SLAs), audits, and compliance certifications that verify the provider’s security practices.
• Data sovereignty considerations: There may be additional assurances about where data is stored and how it is handled to meet legal and regulatory requirements.

Example contexts:

• Cloud storage services that provide encrypted storage with clear privacy policies.
• Managed hosting environments used by financial institutions where data privacy is critical.
• SaaS platforms that operate under rigorous data protection standards.

Open internet-accessible platforms

An open internet-accessible platform is a system or service that is available over the public internet without significant restrictions on user entry or data access. Such platforms are designed to be broadly accessible, often encouraging a wide range of interactions and content sharing among users.

Key characteristics:

• Wide accessibility: Anyone with an internet connection can access the platform, typically without needing specialized credentials or belonging to a closed group.
• Interoperability: Open platforms often support integrations and interactions with other systems or services, fostering an ecosystem of applications and content.
• Transparency and collaboration: These platforms frequently promote open standards and allow users to contribute content, data, or services, which can encourage community-driven development.
• Security considerations: Despite their openness, these platforms often implement security measures (such as user authentication, encryption, and monitoring) to protect against malicious activities while balancing ease of access. However, the protection is often not backed by liability clauses for failure to secure data. 
   

Example contexts:

• Social media networks that allow public profiles and content sharing.
• Public forums or collaborative platforms like open-source project repositories.
• Web services that provide APIs for developers to build upon, such as mapping or weather data services.

Appendix 2: GenAI-relevant legal frameworks and guidelines

This appendix provides an overview of the legal frameworks relevant to GenAI for research in Aotearoa New Zealand.

The information was collated by Dr Simon McCallum using Gemini Advanced 1.5 Pro with Deep Research (19 December 2024). The prompt used was: “When writing policy around generative AI for research in NZ, what are the relevant legal frameworks already in NZ law that need to be taken into consideration?”

International legal frameworks and guidelines

OECD AI Principles: These principles promote responsible stewardship of trustworthy AI that respects human rights and democratic values.
https://www.oecd.org/en/topics/ai-principles.html

UNESCO Recommendation on the Ethics of AI: This recommendation provides a framework for ethical AI development and use, addressing issues such as human rights, diversity, and sustainability. https://www.unesco.org/en/artificial-intelligence/recommendation-ethics

EU AI Act: The EU AI Act is the first comprehensive legal framework on AI, categorising AI systems based on risk and imposing obligations on developers and deployers. While not directly applicable in New Zealand, it may influence future AI regulation in the country.
https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai

Council of Europe AI Convention: This treaty aims to create a legal framework for AI that promotes innovation while managing risks. https://www.privacyworld.blog/2024/09/ai-convention-a-global-framework-for-ai-principles/

United Nations Declaration on the Rights of Indigenous Peoples: This declaration is a comprehensive international human rights document on the rights and freedoms of Indigenous peoples. https://www.un.org/development/desa/indigenouspeoples/wpcontent/uploads/sites/19/2018/11/UNDRIP_E_web.pdf