Who owns the copyright in AI-generated works?
Kiwis love to create, access and share content on social media platforms, streaming services and otherwise over the internet.
Many of those services didn’t even exist when New Zealand’s Copyright Act 1994 (the Act) was last reviewed over a decade ago. Since then, digital developments have created a range of new opportunities to share and access copyright works. Developments in artificial intelligence, data collection, augmented and virtual reality and 3D printing all raise new challenges for copyright law.
This makes the Ministry of Business, Innovation and Employment’s (MBIE) current Issues Paper reviewing copyright law a welcome development. The review is designed to ensure New Zealand’s copyright regime remains robust and flexible in our digital world.
And copyright isn’t only relevant to a select few, like writers and artists. According to Kris Faafoi, the Minister of Commerce and Consumer Affairs, “copyright affects all New Zealanders. We create copyright works when we take a photograph, record a video, or write an email, and we use copyright works by watching a sports broadcast, streaming a movie, listening to music, or reading a book”.
If we want to make the most of new and emerging digital opportunities we need a clear, robust and forward-looking legal framework for the ownership of intellectual property rights. Copyright law is a key part of this picture because it incentivises and protects the commercial exploitation of creative works.
If you have an opinion on how copyright currently operates in New Zealand, then this is your chance to speak up and be heard. While the deadline for submissions on the Issues Paper is 5 April 2019, there will be several more opportunities for public input.
Key issues to consider
- Who owns the copyright in AI-generated works?
As artificial intelligence (AI) becomes increasingly embedded in business operations, the question of who owns the output of AI will become more pressing.
Unlike the copyright laws of most countries, New Zealand’s Copyright Act expressly recognises that copyright works may be “computer-generated”. Despite this forward-looking approach, however, it’s not clear who in practice will own the copyright in an AI-generated work - that is, who has the exclusive right to stop others from commercially exploiting it.
The “author” of a “computer-generated work” is the “person” who “makes the arrangements necessary for the creation of the work” (section 5(2)(a)). But exactly who can be said to have made those “necessary” arrangements? Is it the developer or programmer of the relevant code(s) – and if so, which one(s)? There are likely to be a number of people involved. Is it the customer who is using the AI tool, and who might configure it or add different inputs? For example, a bank configuring a “digital assistant” to answer customer queries specific to its products and services. What about individuals who collect and collate data into meaningful, usable datasets to power the AI tool? Could the AI tool even be interpreted as a “person” in this context?
These questions get even harder in a machine learning (ML) context. ML is a subset of AI that enables algorithms to automatically learn and improve from experience without being explicitly programmed by a human. The Act currently assumes that whenever a “computer” generates a copyright work, a human must have been responsible for making the “arrangements” that enabled that to happen. But while that might have been a valid assumption in 1994 when the Act was last updated, ML techniques now mean there may be no human input beyond the development of the original code. How, then, are we meant to determine whether human “arrangements” were in fact “necessary” for the AI’s creation of the work?
This lack of legal certainty has the potential to inhibit commercial exploitation of AI - why invest in these technologies if your ownership rights are unclear?
- Who owns the data?
We all know that data is “new oil” that greases the digital economy. And there’s plenty of discussion in the business world about who “owns” data – and even who “owns the customer”. Yet New Zealand law is largely silent when it comes to specific enforceable intellectual property rights in data.
The closest we’ve got is copyright in a “compilation”, which could be used to stop third parties copying or distributing a database you’ve created unless they have your permission (see the definition of “literary work” in section 2(1)). But copyright in a compilation will only exist if your database is sufficiently “original” and you can demonstrate you’ve expended sufficient time, skill, labour or investment in producing the compilation.
But what happens when sophisticated software is used to compile information into databases, rather than a human’s “sweat of the brow”? Rather than looking at how much effort and creativity went into compiling a database, is it time to explicitly recognise special rights in data and/or databases under copyright law?
Contracts are often used to plug existing intellectual property law gaps. Contracting parties will often agree who “owns” certain data as between the parties to the contract. But as those rights will generally only apply between the parties to the contract and not further afield, making it difficult to stop random third parties from copying your database.
Ideally the Act would provide greater legal certainty on database ownership to help facilitate the continued growth of New Zealand’s data-driven ecosystem.
- How should taonga works be protected?
Based on Waitangi Tribunal recommendations in the Wai 262 report, MBIE is also seeking submissions on a new, unique regime for “taonga works” or expressions of mātauranga Māori (Māori knowledge).
If implemented, this would be a powerful vehicle for Māori economic self-determination. We suggest that anyone with an interest in the commercialisation of taonga works, such as iwi and Māori trusts, consider making a submission to ensure that appropriate considerations are properly addressed.
Based on the Waitangi Tribunal’s recommendations, MBIE is proposing the creation of a new legal regime that includes:
A prohibition on commercial exploitation of taonga works without first consulting with, or obtaining the consent of, kaitiaki (i.e. individuals with a special fiduciary relationship towards the work);
Allowing such kaitiaki to object to any “derogatory or offensive” use of taonga works; and
Establishing an expert commission to administer those new objection processes, maintain a register of kaitiaki and their taonga works and publish best-practice guidelines for the use, care, protection, and custody of such works.
- Should NZ follow the EU approach?
The European Parliament recently passed significant changes to its copyright laws to “protect creativity in the digital age”.
The European Union Directive on Copyright in the Digital Single Market is designed to limit the sharing of copyright protected content on online platforms. If implemented in its current state, the Direction would impose liability on online platforms for copyright infringement, including for user-generated content. It would also require online news aggregators like Google News to pay publishers for showing snippets of their news stories – the so-called “link tax”.
While the legislation only applies to countries in the European Union (unlike the General Data Protection Regulation, which has extra-territorial effect), it is likely to have a much wider impact. US “Big Tech” companies like Google and Facebook will be affected by the legislation in their European operations.
Should New Zealand follow suit to keep up with global developments or is the European approach a step too far?
Frith Tweedie is digital law leader and Grace Abbott is senior solicitor digital law at EY Law.
Get the latest on digital transformation: Sign up for CIO newsletters for regular updates on CIO news, career tips, views and events. Follow CIO New Zealand on Twitter:@cio_nz
Join the CIO New Zealand group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.