top of page

Authorship; A Case for Non Human Entities (Copyright Law)

Intellectual Property, Copyright, Authorship.


By Chioma Ndu-Akpunku.


Business, Advertising, Marketing, Entrepreneurship.
Oyemaja; Copyright; Authorship, A Case for Non Human Entities.


Authorship is the fulcrum of Copyright. Property rights are centred on ownership and contextually, an author is the owner of the work, thus owner of copyright in the work.


Recent developments have raised questions as to the nexus between ownership and personhood. Should ownership be tied to personhood? Is the ability to create specific to humans? What metrics should be employed for determining authorship where non human entities are involved?


These questions necessitate an inquiry into legislative intent. Legal history is traceable to early civilisation, a concept which is attributed to the activities of man.1 Law aims at curbing human excesses and regulating social relationships, giving credence to the maxim "Law is made for man and not man for law".2 However, beyond social control, law confers rights, a factor particularly relevant in this regard. Laws are made to actualize the ultimate goal of man which is pleasure and the betterment of the society.3


Copyright law fits this category. It stems from the need to promote self fulfilment by recognizing and protecting works of the human mind, while conferring a host of economic rights which provide incentives for these authors. This is achieved through exclusive rights. A key requirement to access these exclusive rights however is authorship.


The term “authorship” in ordinary parlance has no ambiguity. An author is the creator or originator of a work.4 One thing is clear; authorship is a concept reserved for natural persons. The law and interpretation suggests so. Current events however, have stirred debates on the status of non human entities who create original works. The entities most relevant to this discussion are animals and Artificial intelligence (AI).



The Naruto monkey copyright case is a perfect illustration of one of these concerns. In 2011, a monkey 5 took sensational pictures using equipment set up by British Nature photographer David Slater. This resulted in controversy as Slater claimed copyright infringement against the Wikimedia foundation which hosted the image on their page. They alleged in response that animals were not eligible for copyright and the picture was in public domain. Interestingly, the People for the Ethical Treatment of Animals (PETA)6 insisted on Copyright protection for images taken by the monkeys. The US copyright law as well as laws of other nations7 depict copyright in pictures as belonging to the one who takes such pictures. A literal interpretation would suggest the monkey had copyright but courts have repeatedly ruled that animals are not entitled to copyright.8


Artificial intelligence on the other hand utilizes deep learning, natural and machine languages to analyze and interpret data while generating solutions. The uniqueness of AI and it’s applicability in inventions have stirred debates as to whether or not it should be granted copyright. Most arguments in opposition stem from lack of personhood and the fact that AI solutions are the results of datasets generated by humans. Proposing arguments suggest that its learning mechanism gives it sufficient independence and control over output to qualify as self invention. This has been refuted. Recently the British supreme Court held the position that AI does not possess such degree of advancement to self invent.9


AI is not sentient. It requires human intervention to function. From programming to issuing prompts which generate the responses, AI cannot operate alone. This human intervention however bears slight impact on the results. This is the basis of the self invention argument.


Despite arguments to the contrary, it's been revealed over time that AI can create. One recent AI system lending credence to this argument is ChatGPT. Developed by OpenAI, this AI possesses the ability to create literary works based on prompts and a wide range of training data. What this illustrates is that while the input is programmed, the output is not necessarily predictable. A perfect illustration of this creative ability is the Zarya of the dawn dispute which arose out of the registration of a comic book illustrated by the AI platform- Midjourney.10 The US copyright office in February 2023 revoked copyright registration for the images stating that it was not eligible for copyright protection not being a human creation. This can be interpreted as a concession that indeed, AI can create. The implication of refusal of registration however is that those works lapse into public domain. This has consequences. The law generally seeks to protect intellectual property and reward efforts. There are negative implications for the humans whose actions or inactions no matter how little have contributed to such works created by these non human entities (As was the case in the Naruto monkey dispute). This may discourage use of such platforms to invent. Where does this place innovation?


Ammar Reshi’s case beautifully captures AI creation. Reshi, through the implementation of AI systems -Midjourney and ChatGPT created a children’s book titled Alice and Sparkle.11 This sparked intense debate on ethics and AI art. However, there were no doubts as to who the authors were. Midjourney was named author of illustrations while the story was attributed to ChatGPT. This poses serious questions. What level of human involvement determines authorship? Are AI generated works truly original? Ordinarily, qualification for copyright protection entails a fulfilment of certain requirements beyond human authorship - originality and fixation. Originality is mostly at issue in this context.




Oyemaja; Business, Entrepreneurship, leadership
Oyemaja; Copyright; Authorship, A Case for Non Human Entities.



There is little doubt that AI can create. The level of involvement determining authorship will likely be resolved on a case to case basis. However, it suffices to say that sufficient and qualitative contribution on the part of the human would most likely ground human authorship. The words of the court in Feist Publications Inc. V Rural Telephone Service Company Inc.12 are important. Accordingly, an individual claiming authorship for copyright purposes must show "the existence of those facts of originality, of intellectual production, of thought, and conception." Andrien v. Southern Ocean County Chamber of Commerce13, equally attempts to explain authorship. Here, the Third Circuit recognized that "a party can be considered an author when his or her expression of an idea is transposed by mechanical or rote transcription into tangible form under the authority of the party." These definitions embody key elements of authorship.


When applying these principles to works generated by AI on a platform like ChatGPT for instance, one cannot categorically attribute authorship to the human involved. The human would fail the test of thought and conception as these works are not generated through critical thinking but the results of data analysis evoked by prompts.


The originality or otherwise of AI creations cannot be quickly discerned. Ethical concerns have been raised as to the possible derivative nature of AI creations especially AI art. It is deemed that such art is generated from a pool of existing arts in public space, some of which are copyrighted. The same argument is made for literary works. It is arguable however, that the concept of originality is subjective. Originality connotes independence, novelty, creativity, qualities which are neither black nor white. Determining originality in copyright appears challenging as copyright is more concerned with individual expression than novelty. One cannot negate the fact that various factors inspire creativity. People create works under the influence of social, economic and even political motivations. Humans are consciously and unconsciously exposed to these works through social media and everyday experience. This exposure inspires ideas and ultimately result in works which strictly speaking may be derivative. In the same vein, these experiences, facts, information which form tools for humans who create, are compiled and fed into a system for AI to serve the same purpose. It makes for good argument that AI, just like humans, is confronted with series of data made available by social interaction and perhaps, to question AI output would be to question human creations too. How much of our thoughts are independent and not a result of subconscious exposure to certain variables? Questions abound.


Countries have been hesitant towards naming AI as author in copyright and patent registrations. However, recently, the attitude is shifting towards a recognition of the contributions of these machines. The US Copyright office on March 16th 2023 launched new guidelines for the registration of works generated by AI14. It equally launched an initiative to examine law and policy issues raised by Artificial Intelligence. It is expected that these meetings generate positive outcome.


It is apparent that these non human entities lack the capacity to enforce these intellectual property rights hence the insistence on human authorship. It can be said however, that non attribution of intellectual contributions to these entities result in dishonesty and more complications. The right of attribution is one of the fundamental pillars of copyright. The underlying principle is that one who expends physical and mental energy to create should be recognised.


An interesting angle to non human rights is Corporate/juristic Personality. Strictly speaking, a company does not have the physical qualities needed to self execute its objectives but this does not affect its legal personality. A company is also entitled to property rights. It is not out of place to suggest that other non human entities receive juristic personalities in furtherance of certain objectives as well as in recognition of the beneficial roles they play in innovation.


While others have suggested solutions such as an amendment of relevant laws to include the description of computers as authors, the expansion of the work for hire doctrine or to allow such works fall into public domain.15 It is proposed that the joint authorship doctrine be expanded to grant copyright to both the individual and the entity. The result is true attribution of the origin of certain works.




End notes

  1. (Specifically, the activities of the Sumer people of old Mesopotamia).

  2. William Godwin.

  3. This finds roots in the utilitarian theory of Jeremy Bentham which sees the highest good as that which is beneficial to the society as a whole.

  4. Section 51, The Nigerian Copyright Act 1999

  5. A monkey of the Celebes crested macaques species (an endangered species).

  6. An organisation for the protection of Animal rights.

  7. Section 51, The Nigerian Copyright Act 1988.

  8. Naruto v. Slater, No. 16-15469 (9th Cir. 2018)

  9. https://www.economist.com/britain/2023/03/09/can-an-ai-be-an-inventor Can an AI be an inventor? March 9 th 2023.

  10. Wikipedia, “Zarya of the Dawn” https://en.m.wikipedia.org/wiki/Zarya_of_the_Dawn

  11. Wikipedia, “Alice and Sparkle” https://en.m.wikipedia.org/wiki/Alice_and_Sparkle

  12. 499 U.S. 340, 346-347 (1991) (citing Burrow-Giles, 111 U.S. at 59-60). [9]

  13. 927 F.2d 132 (3d Cir. 1991)

  14. https://www.copyright.gov/newsnet/2023/1004.html

  15. Victor, Palace (January 2019). "What if Artificial Intelligence Wrote This? Artificial Intelligence and Copyright Law". Fla. L. Rev. 71 (1): 231–241. https://scholarship.law.ufl.edu/cgi/viewcontent.cgi?article=1439&context=flr




Sources

  1. Megan Mcclintock, “Copyright Laws and AI Writing Bots: How Much of This is Legal?”https://link.medium.com/u8buO1UjJyb

  2. Wikipedia, “Monkey Selfie Copyright dispute” https://en.m.wikipedia.org/wiki/Monkey_selfie_copyright_dispute

  3. CSU Global, “How does AI work?” https://csuglobal.edu/blog/howdoes-ai-actuallywork#:~:text=AI%20systems%20work%20by%20combining,performa nce%20and%20develops%20additional%20expertise.

  4. Harry Guiness, “How does ChatGPT work?” March 31st 2023 https://zapier.com/blog/how-does-chatgpt-work/

  5. Electra Nanou, “Should AI generated art be considered real art? September 17th 2022. https://www.makeuseof.com/is-ai-generatedart-realart/#:~:text=AI%20Art%20Isn't%20Original&text=AI%20generators% 20use%20pre%2Dexisting,a%20way%20never%20encountered%20b efore.



Originally published by Chioma Ndu-Akpunku on LinkedIn

Reach Chioma Ndu-Akpunku on LinkedIn.


Oyemaja Law.

bottom of page