New Zealand’s approach to artificial intelligence (AI) regulation so far has been cautious and pragmatic. Government emphasis has been on flexibility, collaboration, and alignment with international standards. So what steps have actually been taken to date? Sam Irvine investigates for New Zealand Society of Authors and Copyright Licensing New Zealand.
In the middle of last year Cabinet set a strategic direction for New Zealand’s use of AI following the presentation of a paper by then Minister of Science, Innovation and Technology Judith Collins. The paper recommended a ‘light-touch, proportionate and risk-based approach’ to regulation. It proposed leveraging existing principles-based legislation, including the Privacy Act, Human Rights Act, and Fair Trading Act to address AI-related issues rather than enacting standalone legislation. While the Copyright Act clearly has a role to play here, it wasn’t referenced in the paper.[1]
At the same time, Government agreed to draw on a group of intergovernmental principles developed by the OECD to set direction for New Zealand’s approach to AI. New Zealand has long been party to international treaties on copyright and our view at Copyright Licensing New Zealand (CLNZ) is that aligning with neighbours and trading partners on AI regulation makes good sense. While the OECD AI Principles reference respect for human rights and property rights protected by applicable international law, they don’t give specific mention to copyright.[2] After the release, at the same time, of a Public Service AI Framework[3] with principles also informed by those of the OECD, Copyright Licensing New Zealand (CLNZ) responded promptly. We requested that the Government ensure that respect for creative rights, and the risk posed by copyright infringement, are given due consideration in all policy development on AI.
At the end of 2024, we were pleased to be approached by the Ministry of Business, Innovation and Employment (MBIE) a part of consultation on two new pieces of AI policy work. This has been an encouraging step in the right direction. While the work is ongoing and there are no public documents to share yet, our involvement means that conversations around fair reward and copyright are being heard in the wider policy arena.
Though the strategic direction Government has set means there is no immediate plan for comprehensive AI-specific legislation, the Government is developing guidance and frameworks on responsible AI adoption across various sectors. MBIE’s review of the Copyright Act is set to pick up again this year and AI regulation will certainly be on the agenda during that process.
CLNZ will of course be a proactive participant in the Copyright Act review and will also consult closely and get feedback from authors, artists and publishers throughout the process. We’re also currently looking at licence for AI inputs. This would cover AI prompts that reference published work. In the United Kingdom, CLNZ’s counterpart, Copyright Licensing Agency (CLA) has recently extended its licence for corporate and public sector use to include these sort of prompts.[4] Australia are also offering this and CCC in the United States. We hope to have more information on consultation and a New Zealand licence in the coming months. In our conversations with government agencies we continue to highlight the significant role licensing offers in a responsible approach to regulation.
As regulation at home is closely connected to international developments, I’ve included a short summary with updates on AI policy in a few key jurisdictions.
Australia is going through an extensive consultation process and has set up a specific reference group for the creative sector.[5] Though some AI technology companies are advocating for the sort of data mining copyright exception already enacted in Japan, Korea and Singapore for AI, it seems unlikely this will come to pass in Australia. In our conversations with New Zealand government agencies, CLNZ is asking that we stay close to Australian policy development.
In the United Kingdom, in addition to licensing developments initiated by CLA, a rightsholders-led public campaign, Make it Fair, has rallied strongly against the UK government’s proposal on an opt out exception to copyright for AI. Make it Fair has greatly heightened awareness of the issues and the UK government seems to be walking back their position.[6]
Far reaching policy on AI has been enacted by the European Union. The EU AI Act has many responsible elements, including requirements for transparency and privacy, however, deep concerns have been expressed on a recently released third draft of a set of guidelines for compliance with the Act (the General Purpose AI Code of Practice). The International Federation of Reproduction Rights Organisations (IFRRO), of which CLNZ is a member, and a coalition of creative rights holders have penned an open letter saying that the third draft of the Code risks undermining the objectives of the AI Act, infringes European Union law, and misrepresents the intentions of EU lawmakers.[7]
Because policy and regulation, mirroring the tech, is occurring at pace, much has shifted across the last twelve months, in fact the last few months. In all jurisdictions, pressure on AI developers is growing, and more permission-based licensing solutions are being launched. From where I sit, two things are clear: more change is yet to come and, when it comes to AI regulation, including New Zealand’s review of the Copyright Act, there is no need for copyright exceptions. We continue to advocate strongly for permission-based licensing as a fair way to balance AI innovation and the rights of creators in their work.
[1] https://www.mbie.govt.nz/dmsdocument/28913-approach-to-work-on-artificial-intelligence-proactiverelease-pdf
[3] https://www.digital.govt.nz/standards-and-guidance/technology-and-architecture/artificial-intelligence/public-service-artificial-intelligence-framework