Motivation:

This joint investigation presented by the Vidhi Centre for Legal Policy (Vidhi) in collaboration with the Centre for Responsible AI (CeRAI) at IIT Madras explores a participatory model for artificial intelligence (AI) development and governance. As operations get increasingly automated through AI, the various choices and decisions that go into their setup and execution can get transformed, become opaque and obfuscate accountability. This model highlights the importance of involving relevant stakeholders in shaping the design, implementation, and oversight of AI systems. By incorporating input from all relevant parties in decision-making processes, this approach empowers affected users to influence the presence and functioning of AI-based systems, thereby fostering fairer and more accountable outcomes.

The aim of this two part series is to establish the need and importance of a participatory approach to AI Governance while grounding it in real world use cases, through an interdisciplinary collaboration between CeRAI, a premier research centre on responsible AI under the Wadhwani School of DS and AI at IITM, and Vidhi Legal, a leading think-tank on legal and tech policy, between technologists, lawyers and policy researchers.

Paper 1:

In this paper, the authors investigate various issues that have cropped up in the recent past when it comes to AI governance and explore viable solutions. By analyzing how beneficial a participatory approach has been in other domains, they propose a framework that integrates these aspects. The decision sieve, as proposed in the paper, aims to foster participation in every decision making process that culminates in a well rounded AI solution.

Paper 2:

This paper aims to ground the principles established in Paper 1 in the real world. This was done by analysing two use cases of AI solutions and their governance, with one of them being a largely deployed solution in Facial Recognition Technologies which has been widely discussed and well documented, while the other is a possible future application of a relatively newer AI solution in a critical domain. Through these the authors expound and ground the application of the framework proposed in Paper 1.