[SpicyIP] Reflections from an IP Weekend: The First Rajiv K. Luthra Memorial Lecture 2025 Delivered by Prof. Dev Gangjee

4 views

Skip to first unread message

Kartik Sharma

unread,

Dec 1, 2025, 7:08:43 AMDec 1

to spi...@googlegroups.com

The IP enthusiast in me had a great weekend after attending the First Rajiv K. Luthra Memorial Lecture on 29th November 2025, jointly organized by the National Law School of India University (NLSIU), Bengaluru, and the Rajiv K. Luthra Foundation (RKLF). Prof Dev Gangjee from University of Oxford delivered the lecture on the topic ‘Tools or Partners? Hybrid Human-AI Creativity and the Boundaries of Copyright’.

Prof Gangjee was later joined by Eashan Ghosh, In-Charge IPR Chair at National Law University Delhi as the discussant and the session was expertly moderated by Prof Arul George Scaria, NLSIU. It was a delight to hear from this panel since their works and writings have always been a great source of inspiration for me. I am writing this short blog post to share my thoughts and reflections on what I got from the interesting lecture and subsequent discussion.

The theme of the lecture, as evident from its title, pertains to the Gen AI authorship question in Copyright law: Can we claim authorship (and subsequently copyright protection) over works that are a combined product of human intervention and Gen-AI models? Of course the related question of ownership also follows. Prof Gangjee divided his main discussion into three subsets: (i) looking at the journey from human prompts to AI output generally (ii) assessment of situations where the end work arises from AI modifications done to the initial human creation, and (iii) the reverse situation, where Gen-AI creates the initial work and then human modification is performed upon it. Can humans claim copyright protection over works in these different situations?

I won’t summarize everything that Prof Gangjee said, but I will detail a few interesting aspects that stayed with me. He demonstrated a broad contrast between the Chinese position (which is “AI friendly”) and the US law on this (which veers towards protection of creatorship as performed by human creators). The two Chinese cases he discussed (Li v Liu and Feng v. Dongshan Company ) looked at creativity as an iterative and compositional process. As long as the human creator gives detailed prompts successively that fine tunes the desired work through iterations, it would count as sufficient intellectual choices being made as well as the personal expression of the creator would inhere in the work. Here the creator would have in mind a specific vision of the output that they are trying to achieve. This brings associated evidentiary requirements where the creator has to present proof of the process of prompt engineering they used to generate the claimed work. The US law, as it stands (evident most recently through the US Copyright Office report), is highly skeptical towards recognizing human authorship via this prompt engineering. Even if a person puts in detailed prompts, it is the model that carries out the task of creating that work through its specific interpretation of those instructions. In other words, there would be causal uncertainty and a lack of control over the choice of expressive elements in the output. When the human creator decides to go ahead with the AI-generated work, it is them just accepting the specific version generated by the model. The situation would be slightly different if the prompts build on an initial human input. There the report notes that a thin layer of protection limited to the human pictorial authorship will be available, separable from the AI-expression. Sounds like a complicated assessment for adjudicators. I have attached an example from the Report, that Prof Gangjee also referred to.

Image taken from the report.

As I was hearing Prof Gangjee explain all of this and now recollecting it from memory, I find a great deal of food for thought in these opposing approaches to human-AI hybrid creativity.

In the Gen-AI age, how do we then imagine creativity and its value to human society? This is something Prof Gangjee also highlighted when he alluded to the linear versus iterative visions of creativity. Maybe there is something special and unique to human creativity that the law should aim to protect as a policy goal? It is an open question. There is something else also visible in the whole discourse: the tendency to look for inspiration and draw analogies from past developments. There will always be comparisons to how human beings think and create works when the AI generated output will be discussed. Additionally, Mr. Ghosh, in his reflection remarks, aptly noted our tendency to look at history to draw upon for guidance. Photography, an example Prof Gangjee used in his lecture, is one such case study. Photographs captured by human users had initially become the subject of copyrightability debate where similar questions of intellectual effort and labour became prominent. The standing position is that photographers exercise creative choices in deciding the composition, angles, lighting etc, to get to the end result. Cameras, however, have improved over decades and the technological sophistication may have reduced the human input that is needed for a ‘nice picture’. Can photography be a blueprint to decide the big question for Gen-AI? My intuitive response is no, and the readers are most welcome to share their thoughts.

This hybrid Human-AI creativity doesn’t pose just doctrinal queries. We would also be confronted with its societal implications. The proponents of the generative AI revolution tend to align with what is called technopositivism, an almost utopian belief in the potential of technology for societal development. However, Prof Gangjee cautioned that such an ideological framework is often ignorant of questions of politics, inequality and distributive justice. An advancement in gen-AI technology won’t necessarily lead to improvement of the creative culture and enable people to participate in creative production through easier ways. There are no plain causal relationships here. One has to also reckon with possibilities of the detrimental impact of gen-AI outputs on the human artists: will mass produced AI fiction, for instance, substitute and harm the market for existing fiction authors? More fundamentally, how will artists perceive their own creative faculties and their roles in the present age? Will they wholeheartedly adopt technological innovation or resist it?

And as Prof Gangjee rightly mentioned towards his conclusion, generative AI will hold a mirror up to Copyright law where the legal regime will have to answer the lingering uncomfortable questions. It will have to clarify the notions of creativity eligible for protection. Is it the process leading up to the output or the end-result that the law values? The law will also have to think of the implications for the idea-expression dichotomy. And I also think that wrestling with the notions of iterative creativity and causal uncertainty will have some kind of impact on the firewall of aesthetic neutrality that constrains the judiciary’s role in doctrine. The biggest takeaway for me from the lecture was this: satisfactory resolution to many of these issues will have to come from the adoption of an interdisciplinary perspective and a just representation of stakeholder interests in discussions.