The third day of the GPAI Summit hosted a session on AI and Intellectual Property: Protecting Innovation which delved into the dynamic intersection of AI and IP. Recognizing that existing legal structures may not adequately address the unique challenges of AI-driven services, the session explored how legal frameworks can adapt to accommodate this evolving landscape. This could involve topics like data privacy, algorithmic bias, and defining liability in scenarios involving AI. The session likely delved into the intricate nature of IP in the context of AI. This mostly pertained to copyright, patent and trademarks. The discussion also centered around securing datasets and navigating IPR jurisdictions.

The session was graced by the presence of Hon’ble Justice Pratibha Singh of the Delhi High Court who also delivered the keynote address. In her speech, Justice Singh highlighted that by next year the global AI market will be around USD 191 billion whereas in India the number is going to be close to USD 660 million. Speaking about AI and IP intersection, she mentioned that there are 3 broad categories, first is whether AI can be an inventor by itself or with a human being or not, the second is AI as a tool as a catalyst for innovation which she opines is the most realistic possibility and the third is how AI can also be an assistant in IP management, enforcement and protection. She also mentioned that everything that AI is doing today will fall in either of these 3 segments. She spoke about DABUS, the AI generated innovation and how the law is grappling with it. She gave a country-wise comparison on how different countries are grappling with patents when it comes to AI generated content. In the Indian context she mentioned about Section 3(K) which meddles with law related to AI. However, she mentioned that patents are being granted to outcomes where AI is used as an assistant to innovations.

Professor Nick Bostrom, Professor of Applied Ethics, University of Oxford, also delivered a virtual keynote address where he pointed out that AI is extremely fast moving and it will have a transformative impact across all sectors in the coming years. He mentioned about the challenges of patterns, the criteria for predictability. He predicted that “the degree to which AI systems are able to be creative will increase in the coming years.”

Ms Katya Laine, Innovation and Commercialiation WG expert, GPAI, presented two perspectives surrounding AI copyright and regulatory landscape. She spoke on the ethical considerations for granting moral rights to AI generated works. She also spoke about informed consent and user awareness regarding the work involved in the creative process. She also added balancing AI autonomy and human rights which is extremely crucial.

Mr. Barry Scannell, Consultant, William Fry LLP talked about AI outputs i.e., AI creations. He spoke about countries in the EU like France and Italy where copyright works are human-centric. In the US, there is a system of copyrights registration as a result the US has started looking at these issues much before as compared to the rest of the world.

Dr. Maja Bogatog Jancic, Founder & Head, Intellectual Property Institute, who emphasized on data that is required to train algorithms can be protected by copyright. The second thing to consider is the content on which the machine is trained is relevant for copyright. For this, different countries have different opinions.

Ms N. S. Nappinani, Advocate of the Supreme Court and Founder of Cyber Saathi, mostly spoke about deepfakes where she focused on subset of deepfakes i.e., personality rights. The first thing she mentioned was that if deepfakes cannot be a tool for committing crime then where is the good sides and what is role of IP in this. Drawing form examples of her book, she mentioned that deepfakes can have a positive role also. Speaking on fair use of deepfakes she highlighted that “a particular kind of CGI (computer generated image) depending on the mode and manner of the usage is permissible”.

Mr. Pravin Anand, Managing Partner, Anand and Anand Law Firm, quoting an example of deepfakes of the daughter of a celebrity couple stated that when it concerning children there is no excuse for deepfakes. However, he also said that YouTube and Google have no law yet on topics such as this. He mentioned that the patent filing in the country irrespective of the industry has rapidly increased. He opined that software is at the heart of AI as is hardware, datasets and algorithms. Therefore, all the problems that arise for patenting software also arises for AI. He gave the examples of multiple countries including India who have refused to grant patents. Only 2 countries, South Africa and Saudi Arabia, has granted patents.

The session helped in understanding how crucial these complexities are for navigating the ethical and legal landscape of AI development and ensuring fair and sustainable innovation in this rapidly advancing field.