Rudraditya Singh Panwar is a student of Rajiv Gandhi National University of Law & Neer Vaswani is a student of National Law University, Cuttack
I. Generative Artificial Intelligence As A Boon With Complexities
Picture your consciousness housed within a computer’s framework, bestowing upon you command over its functions – that is a true manifestation of artificial intelligence. When this system is given the capacity to self-directedly use available inputs to create original outputs, it signifies the birth of Generative AI. This branch of AI is a transformative power that effortlessly combines creativity and technology to expand the limits of creation. But as AI systems gain more sophisticated abilities to create unique and original content,[1] the existing legal regulations of copyright are being tested to their maximum extent. This prompts us to ponder upon a contentious question of– Who holds the brush and Who owns the painting?
During the 2010s, AI’s limited ability to create small, vague, monochrome visages appeared harmless. However, by 2023, the introduction of technologies like Stable Diffusion, enabled even inexperienced individuals to quickly imitate an artist’s style, leading to a flurry of legal and ethical dilemmas around AI-generated art.[2]
To stir things up, there is no consensus on a precise solution to this dilemma of what works are copyrightable and who is the author of the creation in subject.
II. Artificial Intelligence In Creative Process And Innovation
AI has revolutionized businesses, especially in creative sectors. Yet, it blurs the line between ‘Natural Human Creativity’ and ‘Computer Creativity. For instance “The Road” is a novel created by an AI during a New York to New Orleans road trip. Developed by Ross Goodwin, the AI used data from cameras, GPS, and other sensors to craft its narrative. Despite disjointed prose and errors, it’s praised for its innovative style, termed “pixelated poetry”.[3] Can this creation fit in the traditional definition of creative works in the current age? And can this creativity be something a person can capitalize on?
The rise of AI in creative processes challenges copyright laws that value human creativity. This may lead to “gradable copyrights,” in future where AI collaboration and human ingenuity determine which grade of copyright is granted to a particular creation.
III. Lacunae In The Current Legal Frameworks
It is palpable by multiple court verdicts (UK-Thaler v. Comptroller General of Patents,[4] US-Kris Kashtanova Case)[5] (2022)) around the world that for a work to be copyrighted it has to have a human element, thus works created by human beings by employing generative AI can still be entitled to copyright protection, contingent on the nature of work and the degree of contribution of human creativity (Examples include – AI Generated pictures, Music composed with AI assistance, Design Patterns , Logos etc) .[6]
It is noteworthy in this context to point out that the Hristov (2020) poll, found that half of the participants (57 AI scientists) thought the US Copyright Office was unprepared to handle an inflow of computer-generated works[7].In the Indian context courts[8] strongly assert that artificial intelligence cannot be recognised as the originator of protected works. The Copyright Act 1957[9] mandates that protected works must be the result of the author’s (human’s) expertise, discernment, and inventiveness, which presents a significant obstacle for AI creations since they rely on preexisting data and algorithms established by humans.[10] Section 2(d) defines an “Author” as an individual who creates a work. This definition is explicitly confined to human creators with no reference to AI or autonomous algorithms being recognized as authors.[11] In addition, there have been no policy updates in context of the AI generated content in the Indian framework.
Can the Landmark United States Supreme Court’s 1884 decision in Burrow-Giles Lithographic Co. v. Sarony could be counted as a relevant precedent for future judgments? The rationale discussed in the case concluded that photos can be eligible for copyright protection if the photographer exercises control over the artistic aspects of the image.[12] Does this judgment apply to AI acting as tool of human creativity? To a certain extent, the answer is “No”, as the term “Generative Artificial Intelligence” implies that it operates via its own virtual intellect and has the ability to undertake autonomous processes to make a composition, a capability that a camera lacks.
The diagram depicts different scenarios in which copyright can/cannot be granted , keeping degree of human creativity as the foundation in deciding this. [13][14] (Source of information of diagram taken from United States Copyrights Office)
In a significant ruling by the US Court (Thaler Vs Perlmutter)[15], Thaler was an AI powered machine owner which generated creations autonomously, it was determined that neither the owner of the machine (since the machine is responsible for creating the work) nor the machine itself (as it lacks human qualities) may be granted authorship registration under copyright rules.[16] The author-owner conundrum specifically impacts the software business, leading to ambiguity over programmers not receiving sufficient financial rewards for their development of AI.[17]
A bigger ramification in the software industry comes from artificial intelligence systems, specifically those utilizing neural networks and machine learning, need vast datasets for training purposes, frequently encompassing protected assets like as photographs, videos, and texts. The existing copyright restrictions provide a substantial obstacle to the progress of AI development, since they may prohibit the utilization of these databases for training objectives.[18] This becomes an important consideration when creating new rules since these laws may severely limit the usage of vital information for AI training, such as texts, photographs, and videos, which would hinder the development of AI technology in future.
IV. The Way Forward And Filling The Gaps
Determining copyright ownership in the context of AI-generated works is a complicated matter, similar to choosing who should get credit for creating a cake: the chef who chose the ingredients and set the oven, or the oven itself. In this example, the AI is represented by the oven, and the chef is the human inventor. In most cases, the ‘chef’ has won out in court since the human creator of the AI is acknowledged as the true proprietor of the work. This viewpoint is consistent with a number of court rulings that place a premium on human creativity and oversight. Hence, it is crucial to develop a legislation that explicitly deals with the intricacies of copyright assignment for works assisted by AI, as well as those wholly produced by AI, in turn giving an end to the author-owner dilemma.[19]
Across Europe, including Greece, new developments recommend a unique right for AI-generated art, combining a partial e-moral right for author recognition and a sui generis economic right for financial gains. This new economic right, unlike traditional copyright, ensures fair distribution of profits from AI creations, like sales or licensing revenues. The e-moral right focuses on protecting and acknowledging the human input in AI algorithm development. For instance, a sui generis right might apply to an AI-created painting, ensuring the AI’s developers receive royalties, while the e-moral right ensures their creative contribution is credited. It is important to note that this plan does not give AI a legal personality which can be granted copyrights [20] as it could complicate the value of already granted copyrights and its holder’s rights[21]
A report by ProCreator[22] suggests that approximately 60% of design professionals are currently utilizing AI technologies to augment their creative process thus, in a future characterised by the inevitable presence of artificial intelligence in the creative process[23] a “graded copyright system” presents a pioneering solution. It assesses and assigns a monetary worth to copyrights by considering the contribution of artificial intelligence, effectively balancing the significance of AI’s engagement with the intrinsic worth of human innovation in a particular creation. This framework guarantees equitable recognition and safeguarding the essence of copyrights which is protection of human creativity. Considering the farfetched future, a dedicated copyright tribunal and a centralized AI ledger could be helpful in granting copyrights.
When dealing with the issues presented by AI in the fields of creativity and copyright, it is instructive to reflect on the insights of Sam Hedrick. Hedrick asserts that humans who develop and utilise AI, should maintain enough authority over the AI’s “decisions” to ensure that the use of AI does not impede human ownership of copyrightable computer-generated works.[24]
The intellectual concepts expressed in these works are the product of human creators and controllers of the algorithms (inputs/orders) utilised in the creative process, rather than those of the AI itself” ~ Hedrick. This observation is vital for shaping a legal framework that evolves with the dynamic interplay between human creativity and AI advancements. Recognizing the pivotal role of human intellect in guiding AI, we envision a future where technology harmoniously blends with human innovation, fostering a realm rich in artistic and inventive breakthroughs.
[1] Lv, Z. (2023, January 1). ‘Generative artificial intelligence in the metaverse era’ Cognitive Robotics. https://doi.org/10.1016/j.cogr.2023.06.001 accessed 22 December , 2023.
[2] Vincent, J. (2022, November 15). ‘The scary truth about AI copyright is nobody knows what will happen next’ The Verge. https://www.theverge.com/23444685/generative-ai-copyright-infringement-legal-fair-use-training-data accessed 25 December , 2023.
[3] Goold, P. R. (2020, January 1) ‘Artificial Authors: Case Studies of Copyright in Works of Machine Learning’ Social Science Research Network. https://doi.org/10.2139/ssrn.3734574 accessed 27 December 2023.
[4] Thaler v. Comptroller General of Patents [2023] UKSC 49.
[5] US Copyrights Office, ‘Zarya of the Dawn’ (February 21, 2023) https://copyright.gov/docs/zarya-of-the-dawn.pdf accessed January 8, 2024
[6] Congressional Research Service, ‘Generative Artificial Intelligence and Copyright Law’ (LSB10922, https://crsreports.congress.gov/product/pdf/LSB/LSB10922 accessed 23 December , 2023
[7] Hristov K, ‘Artificial Intelligence and the Copyright Survey’ (2020) Law & Society: Private Law – Intellectual Property eJournal https://doi.org/10.2139/ssrn.3490458 accessed January 9, 2024.
[8] Navigators Logistics Ltd. v. Kashif Qureshi & Ors (2018) 254 DLT 307.
[9] Copyright Act 1957, s-2(d).
[10] Raj, N. (2023, July 28). ‘Legal Implications Of AI-Created Works In India’ Copyright – India. https://www.mondaq.com/india/copyright/1348418/legal-implications-of-ai-created-works-in-india#:~:text=In%20India%2C%20the%20subject%20issue,a%20human%20or%20legal%20person. Accessed 24 December , 2023.
[12] Burrow-Giles Lithographic Co v Sarony (1884) 111 US 53.
[13] ‘Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence.’ (2023, March 16). Federal Register. https://www.federalregister.gov/documents/2023/03/16/2023-05321/copyright-registration-guidance-works-containing-material-generated-by-artificial-intelligence accessed 28 December 2023.
[14] United States Copyright Office, ‘Artificial Intelligence and Copyright’ (Federal Register, Vol. 88, No. 167, 30 August 2023) https://www.copyright.gov/ai/docs/Federal-Register-Document-Artificial-Intelligence-and-Copyright-NOI.pdf. accessed 31 December 2023.
[15] Thaler v Perlmutter (2023) Civil Action No 22-1564 (BAH).
[16] WIPO Lex, United States District Court for the District of Columbia [2023]: Thaler v. Perlmutter, No. 22-CV-384-1564-BAH. (n.d.). https://www.wipo.int/wipolex/en/judgments/details/1840 accessed 1 January 2023.
[17] Lim, Daryl, ‘AI & IP: Innovation & Creativity in an Age of Accelerated Change’ (2018) 52 Akron Law Review 813. https://ssrn.com/abstract=3369200 accessed 5 January 2024.
[18] Mark A Lemley and Bryan Casey, ‘Fair Learning’ (2020) SSRN https://ssrn.com/abstract=3528447 http://dx.doi.org/10.2139/ssrn.3528447. accessed
[19] Y. Burylo, ‘AI Generated Works and Copyright Protection’ (2022) Entrepreneurship, Economy and Law https://doi.org/10.32849/2663-5313/2022.3.01 accessed 8 January 2024.
[20] Erika Hubert, ‘Artificial Intelligence and Copyright Law in a European context’ (Master’s thesis, Lund University,Spring,2020)https://lup.lub.lu.se/luur/download?func=downloadFile&recordOId=9020263&fileOId=9020290 accessed 6 January 2024.
[21] Ring, C. ‘IPO Artificial Intelligence and Intellectual Property: Call for Views Copyright’ (2020) Social Science Research Network https://doi.org/10.2139/SSRN.3763346. accessed 5 January 2024.
[22] ‘Impact Of AI On Design Industry In 2024’ https://www.uxness.in/2023/12/impact-of-ai-on-design-industry-in-2024.html Accessed 8 January 2024.
[23] Anantrasirichai N and Bull D, ‘Artificial Intelligence in the Creative Industries: A Review’ (2021) 55 Artificial Intelligence Review 589 http://dx.doi.org/10.1007/s10462-021-10039-7 accessed 3 January 2024.
[24] Hedrick SF, ‘I Think ; Therefore I Create: Claiming Copyright in the Outputs of Algorithms’ (NYU Journal of Intellectual Property & Entertainment Law, August 21, 2020) https://jipel.law.nyu.edu/vol-8-no-2-1-hedrick/ accessed 10 January 2023.