Governing Artificial Intelligence in India: Data Sourcing, Synthetic Content, and Technological Sovereignty

Artificial intelligence is possibly the most revolutionary technological development of the 21st century. While advances in artificial intelligence have been happening for a long time, it is only recently that it has started interacting with the broader public, triggering a plethora of complex and problematic instances. In this paper, we highlight three concerns, primarily related to data sourcing, synthetic content, and digital sovereignty. The data-related risks are explored in the context of the creative sector, and the concerns of digital sovereignty cut across diverse sectors. The concerns affect diverse policy actors, including AI service providers, users, the creative industry, students, and other communities. To alleviate the risks to these diverse impacted policy actors, we propose a detailed framework for sourcing data for training large language models by focusing on the type of data input and creating a mix of policies by drawing on provisions from the existing Digital Personal Data Protection Act. Finally, when it comes to the indigenisation of technology, we build on the existing IndiaAI mission by highlighting that, in the short run, a public-private partnership model can be sustained, with a focus on collectively bargaining with these technology firms. We believe that these policies are wellsuited to the Indian context, as they build on the existing AI technical and legal policy ecosystem, acknowledging the very requirements and how to navigate them.