Art for Art’s Sake? Analyzing the tryst of AI-generated art & the IP Regime

Art for Art’s Sake? Analyzing the tryst of AI-generated art & the IP Regime

[This piece has been authored by Niranjana Biju, a student at the National University of Advanced Legal Studies, Kochi.]


The brainchild of any entity (emphasis on the term “entity”) may not exist in the form of tangible property, yet it finds itself protected under the broad ambit of intellectual property law. A question may arise as to whether IP law is inherently and essentially anthropocentric. With AI systems causing a creative frenzy, a whole new universe of questions arises regarding this intersection of intellectual property and AI. If your android dreams of electric sheep, who owns the sheep?

Recent news has proven that an AI can be a canvas, a camcorder, and a cello simultaneously. Computer programs today can replicate aspects of creative artistic behavior that were once viewed as exclusive to human ingenuity. This indubitably raises a severe ethical dilemma as to existentially threatening the very concept of artists and creatives, with the possibility of rendering humans obsolete in the creation of art. Though due sympathies are extended to the human parties involved, the essay at hand would like to emphasize exclusively the plethora of legal conundrums raised by AI art.

The very idea of AI-generated art may appear surreal, dystopian even, to the layman, causing reasonable apprehension on their parts as to accepting the same. This may assume that creativity is inherently human; this could be attributed to the general use of vague terms such as “inspiration” to explain the origin of creative thought. The idea of a creative thought emerging from a software algorithm thus seems far-fetched to the common man. Nevertheless, the fact that humans are not conscious of how a creative idea manifests itself does not necessarily imply that a scientific explanation cannot exist.


While introspecting upon the intersection of AI art and IP law, two major questions arise:      can an artwork generated by AI be protected under Copyright law? If yes, who would be entitled to the ownership rights emerging from said copyrighted work? Secondly, does the very process of generating AI art infringe the copyright of existing artworks, rendering it purely derivative? While discussing the possibilities of conferring copyright protection to AI-generated artwork, there arises the question of granting legal status to AI, which would be nothing short of opening a can of worms.


Assuming that AI-generated work can indeed receive copyright, there exist various approaches to ascertaining its ownership. One approach would be to confer ownership rights to the software developer. Technically speaking, there would not be a final product to be considered for copyright protection had there not been an AI system which is a result of the sweat and tears of the developer. Providing ownership to the copyright and thus allowing for royalties of the proceeds from the AI would incentivize and further enable the developer to continue with their work.[1] Though this may appear to be the only suitable option, critics have raised valid points to the contrary. Firstly, the payment made to the developer for using the AI tool should be enough of an incentive. More importantly, the developer’s role is merely providing the necessary tool, which is utilized by the user at their       discretion leading to the output in which the developer shares zero stakes.

An extension of this criticism would call for ownership rights to be extended to the user of the AI tool. Considering that the users have paid to use certain software of the developer and made significant decisions without which the program would have remained impractical, the user shall be entitled to ownership of the copyright. This practice would be an application of the Made for Hire doctrine. However, it is to be noted that not every AI tool requires the same amount of human contribution as the GANbreeder.

A middle-ground here would be to allow for joint ownership of the AI-generated artwork, as in the case of RAGHAV. Though this may appear to be a clear-cut solution, it has its own set of challenges. In a situation where there exist multiple parties with varying levels of contribution, the task of accrediting each party in the absence of a quantifying system would be highly subjective. This arrangement also provides for double profits to a certain party which is highly unfair to the others involved. Now, other than allowing the artwork to exist in the public domain, which will eventually disincentivize users from displaying their art, the only viable option is to confer the AI itself with the ownership of its artwork.


The most conceptually intriguing yet non-feasible option would be to confer ownership rights of the copyright to the AI system itself. For any entity to have a copyright under its name, it should have a legal personality. Now, the idea of AI being entitled to a legal status comparable to natural persons will cause a paradigm shift in the field of IP law with far-reaching implications. To answer whether a particular AI system is entitled to legal status is complex, for it largely depends on the relevant jurisdiction and subsequent application of the appropriate doctrine. The Sweat of the Brow doctrine. provides copyright protection on the basis of the labor, skill, and investment of capital put in by the creator. One could       argue that AI systems do employ appropriate skills and undertake intensive labor in the generation of art. Now, the Modicum of Creativity doctrine requires the art in question not only to be the product of independent creation but also to exhibit a minimal degree of creativity. The idea of AI being creative, as discussed earlier, may not be palatable to society at large, posing more of a moral than a factual dilemma with greater social implications. The lack of intentionality involved in the generation of the art, following Searle’s Chinese Room argument, and the lack of conscious thinking of the AI system may make it appear as only “apparently creative.” Indian laws have been silent on the whole AI art debacle, with no policy guidance from the Indian Copyright Office as to whether only humans can be considered authors. Nevertheless, it has been made apparent that it favors the requirement of a “flavor of minimum requirement of creativity.”


Coming back to the second question, it is to be noted that AI’s method of collection of information may be highly susceptible to causing copyright infringement. This is because machine-learning software utilizes a database of work collated from other artists. Essentially, through the cumulative acquisition of knowledge, the AI generates artwork. The creation of said artwork involves following certain rules to ensure that that artwork does appear odd to the spectators. The AI simply cannot break these rules, for it is incapable of imagining a visual beyond what it has been taught. Thus, there exist limitations on its creativity, unlike that of humans. All of this leads to a heavy potential for the generation of an unauthorized derivative in the process of reviewing existing artwork.

The birth of any brainchild is a result of the heavy influence of its predecessors. It is common knowledge that no creative work arises from a vacuum; it is a fruit of cultural inheritance and lived experiences. Thus, nothing can ever be truly original, and to expect the same from AI-generated art may appear discriminatory. Nevertheless, it would be wise for persons utilizing AI systems to remain vigilant and ensure that their artwork contains a distinct mark       to avoid creating a substantial likelihood of confusion and further lawsuits. Also, the geniuses behind AI technology need to ensure they are complying with copyright laws to prevent infringing on existing artists’ work with their datasets.


Currently, very few legal precedents exist that were drafted with AI-generated art in mind. This essentially means that AI-generated art is out in the cold and calls for strategic and creative solutions to tackle the numerous issues posed by the same. With new advances in technology often being accompanied by unforeseen and even unwelcome legal consequences, there lies an air of uncertainty that is to be duly navigated by concerned IP law regimes. This legal uncertainty in an ever-evolving landscape makes it hard for brands and creatives to strategize and avoid unforeseen legal repercussions. For now, artists can rest easier knowing there’s still a place, demand, and protection for human creativity.




Leave a Reply

Your email address will not be published. Required fields are marked *