Let’s IPsa Loquitur: Dr. Zakir Thomas on Copyright, AI, Technology Sovereignty and Open Access
Sonisha Srinivasan
In our 2nd episode of Let’s IPsa Loquitur, we present to you an interview with Dr. Zakir Thomas on Copyright, AI, Technology Sovereignty and Open Access. A retired IRS officer, Dr Thomas is also a former Copyright Registrar, as well as a former Project Director of CSIR’s Open Source Drug Discovery Project. (Full bio below)
I met Dr. Zakir Thomas at the SpicyIP Summer School, where he was one of the organisers. He added immense value to the learning process by actively breaking down concepts after certain sessions, providing thought-provoking insights such as the rationale behind the current copyright term, and constantly engaging with students to gauge the level of understanding and improve the learning process. In the backdrop of those encouraging and insightful interactions, it was such a pleasure to have this conversation with him. In this episode, I speak to him about what appears to be the hot topic of the year: the numerous legal and policy issues around Generative AI and Copyright. Below, I’ve outlined some key points of what we covered in our discussion, along with the reading materials you can refer to. (Please note: this was recorded in September, so some connected developments that have happened since have naturally not been touched upon.)
We started with some basics. Dr. Zakir Thomas breaks down the process of how Large Language Models (LLMs) work, and we discuss the stance of different stakeholders to try and uncover where interests actually lie. We also touched upon Text and Data Mining (you can read more about TDM here, here, and here.)
On the development of regulations around AI, Dr. Thomas drew an analogy with the development of regulations on the internet. He highlighted that while WIPO quickly came up with the Internet treaties- WIPO Copyright Treaty (WCT) and WIPO Performances and Phonograms Treaty(WPPT) within just a few years of the Internet coming into being, the organization is still in the initial stages of learning when it comes to AI technology. He considers this an opportunity for students of law to engage with.
Don’t miss out on his remarks on the “Black box theory”, as well as how “memorization” really works while training data. In this regard, reference is made to the dispute between The New York Times and OpenAI (you can read about them here and here). Some other interesting highlights include the numerous issues beyond Copyright Law which are relevant and important in this discussion of AI training, emphasis on training foundational models in regional languages in a manner to reflect the diverse country’s culture, belief, and ethos, the discussion on Technology Sovereignty, as well as the role of Open Source Movement in reaching a certain level of autonomy.
A small flashback to the pre-AI era reminds us of how the Copyright Amendment Act 2012 was seen to be a solution to bargaining power concentrated in the hands of music publishers and labels, while authors and composers of the underlying work were being denied their share of royalties, in the context of the music industry. (Interested readers can refer to Prashant Reddy T.’s posts here, here, here, and his long paper on the 2012 amendments here.) Fast forward to the post-AI era, in the context of how AI companies have become a repository of knowledge, acting as an intermediary for the public to produce and consume information, Dr. Thomas brings about a reference to the Sony Betamax case.
In the context of an analysis of a paper by Benjamin Sobel titled “Copyright Accelerationism” (a brief summary here in the blog), as well as the recently imposed ban by the Delhi High Court on shadow libraries like Sci Hub, we discuss the implications of this ban. Dr. Thomas sees the solution to the Sci Hub ban in schemes such as O.N.O.S. (You can read more about it here and here). He discusses the challenges to Open Access to Information, where technological barriers prevent access.
You can scroll down to watch the full episode.
You can follow SpicyIP TV to watch the first episode of Let’s IPsa Loquitur, an immersive discussion on personality rights, live sessions on the authority of CCI to hear anti-competitive patent complaints, and on higher education in IP, as well as the ongoing Summer School series.
Let’s IPsa Loquitur: Dr. Zakir Thomas on Copyright, AI, Technology Sovereignty and Open Access
You can watch the video here: https://youtu.be/ZW_ZO6qQgo0?si=uvtMcI6vR7A-nW7I
About the episode:
Sonisha Srinivasan speaks to Dr. Zakir Thomas about the intersection of Generative AI with Copyright law and Policy, Technology Sovereignty, Open Access to information and the broader regulatory challenges shaping the future of AI.
About Dr. Zakir Thomas:
Dr. Zakir Thomas has 35 years of experience of working in the Government of India as an officer of Indian Revenue Service had been instrumental in some of the major policy initiatives in field of intellectual property as a former Registrar of Copyrights and the Project Director of Open Source Drug Discovery Project. He holds an MSc (Physics), LLB (Delhi University) and Master of Intellectual Property, Commerce and Technology from Franklin Pierce Law Centre, University of New Hampshire, United States. He earned his PhD from the School of Social Sciences IGNOU, Delhi. He has authored several book chapters and articles in peer reviewed journals on Intellectual Property law and its interplay with open source innovation, pharmaceutical industry and copyright law. He was an invited faculty in WIPO seminars on Intellectual Property, delivered talks in the European Union Parliament, American Chemical Society and several other international fora. He was a faculty for a two credit course in NALSAR, Hyderabad. He has given talks in leading national institutions on copyright law, patent law and innovation. He has been the Indian delegate at WIPO and at the OECD tax forum. Dr Thomas is recipient of the Finance Ministers Award (2017), the highest honour conferred by the Income Tax department. Prior to retirement, he was Director General (Systems) of the Income Tax department, leading one of the largest e-governance projects in the country.
All views expressed here by the speakers are in their personal capacity, and in general are to be considered as curious engagement with the subject at hand. If you want the formal language, then: This podcast is intended solely for informational and academic discussion purposes and does not constitute legal advice or create any attorney-client relationship.
List of Additional Readings:-
- Goutham Rajeev and Vedant Bharadwaj Singh, Training GenAI: Infringement or Fair Use? (here)
- Varsha Jhawar, An IP-Centric Approach towards AI Regulation in India- Part II (here)
- US President Trump’s Executive Order, Removing Barriers To American Leadership In Artificial Intelligence (here)
- Aman Gupta, Apple faces copyright lawsuit over training Apple Intelligence AI on pirated books: What we know so far (here)Jean Gil Barroca and Alfons Buxo, Cloud sovereignty: Three imperatives for the European public sector (here)
- Akshat Agrawal, ANI Media v. Open AI : The Opt-Out Strategy (here)
- Akshat Agrawal and Sneha Jain, Indian Copyright Law and Generative AI (here)
- Shivam Kaushik, Part I- Applying Natural Intelligence (NI) to Artificial Intelligence (AI): Understanding ‘Why’ Training ChatGPT Transcends the Contours of Copyright (here)
- Shivam Kaushik, Part II- Applying Natural Intelligence (NI) to Artificial Intelligence (AI): Understanding ‘Why’ Training ChatGPT Transcends the Contours of Copyright (here)
- Bharathwaj Ramakrishnan, Taking Stock of ANI vs OpenAI Copyright Litigation- Part II (here)
- Shama Mahajan, [Part I] ANI v. Open AI – A Lesson in Resisting the Temptation to Borrow Excessively without Legislative Sanction (here)
- Shama Mahajan, [Part II] ANI v. Open AI – The Storage Paradox is More Than Just Transient! (here)