18 January 2021
Copyright and Artificial Intelligence
By Faye Cheung, Researcher
The Intellectual Property Office recently called for views regarding the relationship between artificial intelligence (AI) and copyright. The consultation addresses an apparent anxiety over whether the Copyright Designs and Patents Act (1988) is keeping up with developments in AI and machine learning. The increase in the use of AI and it’s ability to generate works in which copyright could subsist, such as music, visual art or literature, raises questions regarding what might be deemed as non-human authorship. An example is the Portrait of Edmond Belamy, which sold for $432,500 at Christie’s New York auction house in 2018, for more than 40 times its $10,000 estimate. The portrait was created by an algorithm and was the world’s first artwork created by artificial intelligence to be sold at auction.
The IPO’s call for views referred to the Government definition of AI, which is ‘technologies with the ability to perform tasks that would otherwise require human intelligence. The US Copyright Office do not currently register works that have non-human authorship, for example, works produced by ‘nature, animals, or plants’ or works ‘purportedly created by divine or supernatural beings’! They also do not have a provision for AI. However, the UK’s copyright legislation acknowledges that works can be created by machines:
“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”
This provision is developed in Section 178 of the Copyright, Designs and Patents Act 1988, which clarifies that computer-generated works occur when there is ‘no human author of the work’.
Despite having a well-established legal framework for computer-generated works there is ongoing debate regarding the copyright status of work generated by artificial intelligence. This is because the creative outputs of programmes that use artificial intelligence and machine learning largely depend on the reading and processing of work created by others.
The British Copyright Council (BCC) recently published its response to the Intellectual Property Office’s calls for views. They argue that the copyright status of works used to ‘inform and establish value within the AI applications’ should be respected and that no legal exemptions should be made for AI developers. There is a concern that exemptions will be made for AI developers such as the exemption that was introduced for text and data mining (TDM). Instead of an exemption or a derogation on existing copyright law, which could see existing copyright infringed, the BCC argues that the solution lies in licensing source materials and in the existing ownership structures.
The IPO’s consultation also covers patents, designs, trademarks and trade secrets. All of which are areas that will be hugely impact by AI and machine learning, which may be used to create new works or to help with the monitoring or enforcement of IP rights. The IPO are currently analysing all feedback from the consultation and the outcome will be published on the IPO website in due course. We watch these developments with great interest.
 United States Copyright Office, Compendium of U.S. Copyright Office Practices (third edn, US Copyright Office, December 22 2014) [313.2]
 Copyright, Designs and Patents Act 1988, Section 9 (3)
© Naomi Korn Associates, 2021. Some Rights Reserved. The text is licensed for use under a Creative Commons Attribution Share Alike Licence (CC BY SA)
Disclaimer: The material in this blog post is for general information only and is not legal advice. Always consult a qualified lawyer about a specific legal problem.