“Closed-door approach” of AI poses concerns in terms of fairness/bias issues, inaccuracies/glitches, privacy/security issues and explainability/transparency issues. Deploying AI systems and services to crucial sectors with potentially large social impact needs to be built and deployed such that there be no harms to the users and stakeholders. At the same time it is also important there be adequate measures to voice concerns and feedback from the stakeholders/users and it is even more a challenging task to monitor these concerns and incorporate the necessary changes to the AI systems.
This project aims to come up with a novel policy framework – “Participatory AI”, where multiple stakeholders involved in the usage of AI systems/services can come together and contribute to the design and development of the AI system such that it will be impactful to the society in a positive manner.
This project is a crucial step towards democratizing AI and making AI deployability safe, responsible and reliable. Vidhi Centre for Legal Policy, an think tank on public and legal policies, is partnering with CeRAI.