Introduction
Artificial Intelligence can generate written and visual content in seconds. This content can be used for various purposes such as marketing, presentations, advice, websites etc. From a legal standpoint, the important question remains who holds the intellectual property rights to this content and what are the current copyright laws? Theoretically, it could be the individual who prompted the response, the AI itself or the organisation that trained the AI. This is important to deter plagiarism and content theft. I believe that the individual who prompted the AI to produce the content should be seen as the owner, author or artist. This view seems far more logical than the AI or company behind it having these rights. The legal framework surrounding this needs to be clear and thorough to safeguard human creativity and ownership. The following article will dive deeper into the current legislation and case law surrounding AI content rights in the UK. My initial hypothesis is that the answer to this question will remain open-ended with a need for more privacy and IP regulation in future.
Legislation
The starting point is the Copyright, Design and Patents Act 1988 (CDPA). Section 1 states that copyright is a right relating to CDPA section 1;
(1)(a): original literacy, dramatic, musical or artistic works,
(1)(b): sound recordings, films (or broadcasts) and
(1)(c): published editions.
Furthermore, for copyright protection to apply, the requirements in section 153 must be met. Section 153 sets out certain requirements and refers to other sections in the process. For example, section 153 refers to section 154, which requires that the author is a British citizen, national or British protected person, UK resident or body incorporated under UK law. The initial requirements already raise questions regarding how AI fits into these. Section 2 of CDPA sets out what rights apply to protected works. Section 3 to section 8 set out descriptions of the works set out in section 1 above. Section 9 is of great importance to the above questions. It describes the author of the work as the person who created it. Section 9(3) goes further and states that the author of literary, dramatic, musical or artistic work which has been computer-generated is the person who made the necessary arrangements for the creation of the work. Section 178 defines computer generated as work created by a computer with no human author. This favours the logical view described above that the person who prompted the AI content could be making the necessary arrangements but it could also be the AI developer who trained the AI. Section 9(3) deals with who the author of computer-generated content is but does not assist with whether copyright is original and afforded protection[1]. A further issue arises by virtue of section 79(2)(c) depriving the author of the right to be identified as the author if the work is computer generated. [2]
Therefore, although the CDPA recognises computer-generated content, it is unclear who the person who made the necessary arrangements is. The act also states that this person is the author but deprives them of being identified as the author. These regulations need to be clearer and amended to deal with AI-generated content as opposed to computer-generated. Furthermore the act refers to a person, which rules out the AI being the author. However, AI bots or robots may be granted some kind of legal personality in future, in the same way, a company has a separate legal personality. This logical legal development would allow AI-powered assistants to act as agents for individuals and enter contracts on their behalf.
Conclusion
The question of originality is also not clear-cut. The UK test for originality was expressed in University of London Press Ltd v University Tutorial Press Ltd [3]. It was described as whether the work is an expression of the author's skill, labour or judgment.[4] It is still unclear how this test will be applied to the level of originality of prompted AI content. On the one hand, basic prompts can still generate a large quantity of content, which will certainly lack originality. However, prompting can be complex and require a strong level of skill and judgment. We will have to wait for a decision from the Supreme Court on this issue. Furthermore, AI may be integrated into platforms which have their user terms, these terms might state that the user of the platform is the owner of AI output. This would not align well with section 9(3) of the CDPA and the meaning of necessary arrangements. In conclusion, there is a lot of room for new legislation and regulations which will hopefully ensure the safety and security of human individualism and originality.
Footnotes
[1] J Dickenson, A Morgan & B Clark, Creative machines: ownership of copyright in content created by artificial intelligence applications (2017) 39(8) E.I..P.R pp 457-460.
[2] J Dickenson, creative machine E.I.P.R 458.
[3] (1916) 2 CH. 601.
[4] J Dickenson, creative machine E.I.P.R 459.
Comments