Copyright law is being challenged by disruptive technologies such as AI and blockchain, themes addressed in the recently published 5th edition of the author’s book Digital Copyright, on which this article is based.
Artificial intelligence and virtual reality
The rise of artificial intelligence (AI) and virtual reality (VR) are starting to pose significant challenges to how we conceive and apply copyright – whilst AI and VR are distinct concepts they are linked together in this discussion as they raise related issues in terms of authorship and protection.
Copyright law is rooted in two significant assumptions. Firstly, that there is an author in relation to the work – even if the work is computer-generated there is still deemed to be an “author” – “the person by whom the arrangements necessary for the creation of the work are undertaken” (section 9(3) CDPA). As has recently been commented, “[m]ost jurisdictions, including Spain and Germany, state that only works created by a human can be protected by copyright” (Guadamuz, Artificial Intelligence and Copyright). However, the UK acknowledges that a computer-generated work is a work generated by a computer in circumstances such that there is no human author of the work (section 178 CDPA) and this can can still be protected by copyright (section 9(3) CDPA).
Secondly, in UK law at least, for a work to be protected by copyright it has to fall into one of the statutory definitions of a copyright work – a literary work, artistic work and so on. This application of categories continues to raise issues for works which cannot be neatly categorised, although attempts are made to do so, for example in discussing virtual works two recent scholars note that virtual works can include: computer code that creates virtual environments, the audiovisual presentation to the user, the interactive media including tactile components of the environment which are experienced by the user, as well as video recordings of audiovisual components to be played in goggles.
AI will clearly generate new copyright works – for the moment at least the current law will continue to apply, which means that issues of ownership and duration of any copyright works created will remain to be governed by sections 9(3) and 12(7) CDPA. The interaction of human beings with machines to create VR environments also raises issues around joint authorship in certain cases, as well as other legal issues.
The current law on computer-generated works dates back to viewing computers as simply human tools; machine learning and other powerful new AI techniques challenge this assumption. As Guadamuz has noted, “[W]ith the latest types of artificial intelligence, the computer program is no longer a tool; it actually makes many of the decisions involved in the creative process without human intervention”. Unless computers eventually become “persons” in their own right, for now at least there are two options in applying copyright law to computer-generated works – either there is no copyright protection at all (because there is no human author) or the UK approach is followed and the programmer becomes the author. Certainly the UK approach is pragmatic as it protects the investment companies make in AI. However, this approach is still problematic and somewhat simplistic – what account is taken of how the user might interact with the program and the IPR that may subsist in the data used to train the algorithms used?
Robonotices and algorithmic enforcement
Copyright ought to provide for a balanced community of interests – authors/owners of copyrights have rights but these are balanced by rights of fair use/fair dealing granted to copyright users. Yet digital copyright has the ability to shift this balance in favour of authors/owners. It has also introduced a new class of copyright actor – the online intermediary or platform – Google, Facebook and so on – who make available content posted or created by users and have the power to take it down as readily as putting it up in the first place. From the 1990s onwards online intermediaries have had a privileged status in copyright law – if they make available infringing content then as long as they take it down once they become aware it is infringing (eg by receiving a notice and take down request) then they have a “safe harbour” from copyright infringement. What has happened in recent years is that the whole process of monitoring and enforcing online copyrights has become automated:
“[M]ajor copyright owners increasingly exploit robots to send immense volumes of takedown requests and major online intermediaries, in response, use algorithms to filter, block, and disable access to allegedly infringing content automatically, with little or no human intervention” (Perel and Elkin-Koren in (2016) 19 Stan Tech L Rev 473).
Such “algorithmic law enforcement”, it has been argued, allows little consideration for fair use and effectively puts online access to content in the hands of online intermediaries and represents a fundamental shift in copyright governance. Yet it seems here to stay despite concerns that such automated anti-piracy systems (AAPSs) can lead to the over-blocking of content with no balancing mechanism and the endangerment of cultural diversity in the long run (the conclusion of Jacques et al in (2018) 40(4) EIPR 218–29).
Blockchain and copyright
Blockchain is a rapidly developing technology that can best be described as a form of digital ledger that is shared and distributed (so that each participant has a duplicate copy) – it can facilitate the recording of transactions and the tracking of assets. The ledger is secured through the use of cryptography and is immutable. Blockchain owes its name to how it stores transaction data – in blocks that are linked together to form a chain. Blockchain can also be used to develop “smart” contracts, ie an agreement or set of rules that apply to a transaction – it is stored on the blockchain and is executed automatically as part of a transaction. For example a smart contract might contain a set of rules for how a digital copyright asset such as a digital photograph is licensed. Bitcoin is also an example of the use of blockchain. Blockchains can be public or private.
Because blockchains are immutable ledgers of data they have an obvious use in rights management. Metadata on ownership and other aspects of the digital copyright asset can be stored on the blockchain. A smart contract can be used to automate who has access to the asset and under what conditions. The blockchain platform can also handle remuneration through for example the use of a cryptocurrency like bitcoin. It therefore has the potential to revolutionise how digital copyright assets are exploited – an example of this is the KODAKOne rights management platform for digital images and there are a number of blockchain initiatives in the film and music industries as well, for example Jaak which is developing a blockchain based licensing and content distribution environment. The UK copyright collecting society DACS has also been exploring the use of blockchain.
It has also been suggested that copyright could become redundant if content is managed through blockchain and smart contracts. There would be no need for it although there must remain a risk of piracy. One might argue that blockchain is simply a more fancy form of digital rights management (DRM) which has been around for many years. Certainly if blockchain technologies are used to publish and distribute content there will be issues around how copyright limitations and exceptions might apply, the exhaustion of rights in the content and so on. In the CDPA a number of copyright exceptions cannot be overridden by contract. So, as an example, the research and private study “fair dealing” exception cannot be restricted by contract: “To the extent that a term of a contract purports to prevent or restrict the doing of any act which, by virtue of this section, would not infringe copyright, that term is unenforceable” (s 29(4B) CDPA). “Contract” is not defined so arguably it could include a smart contract but in practical terms if a rights management platform were to technologically prevent a user availing themselves of a copyright exception it is difficult to see what effective remedies a user might have under current law.
Another issue blockchain raises is that the autonomous, resilient and tamper-resistent features of blockchain pose challenges if blockchain is used for unlawful purposes such as illicit file-sharing. Blockchain removes the need for an intermediary so there is no scope for the notice and take down regime of the US DMCA and the Electronic Commerce Directive to apply – indeed the immutable nature of blockchain can render take down impossible. In practice the only way to prevent the operation of such blockchain based systems is for content owners to target individual users/infringers which in a large scale system may be impossible – effectively authors’ rights are undermined because they cannot be effectively enforced (see De Filippi and Wright, Blockchain and the Law, pp 123–25).
Simon Stokes is a partner with Blake Morgan and leads their London technology team. His copyright practice spans software, databases, financial technology, images and publishing. Email email@example.com.
Image: Adapted detail from the cover of Digital Copyright.
Digital Copyright: Law and Practice 5th edition by Simon Stokes, published January 2019 by Hart Publishing. Hardback 360 pp, £60+p&p. Ebook £64.80 inc VAT. The book helps put digital copyright law and policy into perspective and provides practical guidance for those creating or exploiting digital content or technology. Further details.