What’s also unclear is the question of authorship. This is especially true for the creative works based on machine learning. Since the input data allows AI to learn from is made by someone else, if the outcome is applied for commercial use, should it not infringe the copyright of the owners of those provided data? Obviously, the owners can’t sue software, people who use a tool that caused infringement on others works would be accused. Could someone with a significant influence on the sound of modern music—for instance, the Beatles—argue that their copyrights are infringed by software designed to deconstruct, analyze, and iterate on their actual work? Songwriting licensing will be the hard problem to handle those pieces of commercially released music intending to emulate, but not duplicate, another artist in the litigious world. 

According to the current state of the law, creative works qualify for copyright protection only if they are original. Most jurisdiction definitions of originality, including Spain and Germany, state that copyright is only conferred to works created by a human author. In the United States, though there are no indications that law to the ownership of non-human generated works would be amendable, the law has recently clarified the qualifications in the copyright of such work. In 2016, in response to a United States court ruling in a copyright infringement case involving a monkey who had taken a selfie using a camera that a British photographer had set up, the U.S. Copyright Office updated its rules to specify that copyright law only accept the registration for “the fruits of intellectual labor” that “are founded in the creative powers of the mind”’. In section 313.2 of the most recent edition of the Compendium of the U.S. Copyright Office Practices, Third Edition, its notes that “to qualify as a work of ‘authorship’ a work must be created by a human being.” The rules listed important specific examples include works “produced by a machine or mere mechanical process that operates randomly or automatically without any creative input or intervention from a human author” are ineligible.

But there are still many emerging problems about copyright infringement and artificial intelligence in the new business model are waiting for addressing. For machine learning software, a computer program is no longer a tool, once humans program the well-ruled algorithms, then the decision making – if we considered it as the creative spark – during creative process comes almost entirely from the machine, and the creative work can be produced actually without human intervention. According to the law, the artificial intellectual property is not eligible for the copyright protection because of the lack of “intellectual creation from the human author”. For the companies offering the products and services base on AI systems or consultancy, like AI-generated music, it would make a huge impact on the creative economy if their merchandises would be used without payment by anyone in the world.

There is another provision of governing intellectual property ownership and exploitation in some countries such as UK, Hong Kong, India, New Zealand and Ireland. In the UK copyright law, which is best encapsulated this provision, states In the case of a literary, dramatic, musical or artistic work which is computer-generated, the author shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken.” in section 9(3) of the Copyright, Designs and Patents Act 1988 (CDPA). In addition, section 178 of the CDPA defines a computer-generated work as one that “is generated by computer in circumstances such that there is no human author of the work”. It leaves arguably room for who the law would consider being the person making the arrangements necessary for the creation to be generated.

Take the case of Microsoft Word(Wipo.int., 2017). Numerous people use Word to produce their literary text works, apparently the computer program does not own every piece of works, the copyright belongs to the author who runs the program to create his or her work. But when it comes to artificial intelligence algorithms that are capable of generating a work, the contribution of the human user to the creative process may simply be to press a button to let the machine do the things left. This question could be solved by looking at works on a case by case basis. In the English case Nova Productions’ copyright dispute with Mazooma Games and Bell Fruit Games (EWCA Civ 219, 2007), the Court of Appeal have to judge on the authorship of a computer game, and highlights a player’s input “is not artistic in nature and he has contributed no skill or labour of an artistic kind”. So considering user action case by case could be one possible solution to the problem.

At this stage, copyright law is still grappling with the authorship of artificial intelligence-generated content. Although it has been moving away from originality standards that reward skill, labour and effort. It is the foreseeable future that more copyright protection exceptions will be established when it comes to the results of emergent works created by intelligent algorithms with little or no human intervention. This will ensure companies keep investing in the technology, safe in the knowledge they will reap the benefits. One day it might hold the debate about whether AI can be the creator legally, do AI itself need copyright protection and should be given the status and rights alike the human being? It will be an interesting question to think of, but is totally a whole other story.



Wipo.int. (2017). Artificial intelligence and copyright. Available at: http://www.wipo.int/wipo_magazine/en/2017/05/article_0003.html (Accessed 23 Nov. 2017).
Nova Productions Ltd v Mazooma Games Ltd [2007] EWCA Civ 219.