Google invests $1 mn in IIT Madras’ new Centre for Responsible AI

The announcement came during the Centre’s inaugural workshop and panel discussion held on Monday. The AI division of IIT Madras is slated to support research projects and develop datasets for AI applications.

CeRAI was formally inaugurated on April 27, with the ceremony presided over by Rajeev Chandrasekhar, the Union Minister of State for Electronics and Information Technology.

The Centre has forged collaborations with the industry body Nasscom, Southern Indian Chambers of Commerce and Industry (SICCI), the policy thinktank Vidhi Centre for Legal Policy, and the Ministry of External Affairs (MEA)-affiliated thinktank Research and Information Systems (RIS).

These partnerships aim to promote a responsible use of AI through academic curriculum development, exploring implications of AI, developing a participative AI framework, and mentoring startups to create responsible AI applications. The startup mentorship and incubation organization, The Indus Entrepreneurs (TIE), will also be associated with CeRAI, as stated in a press release.

As part of its AI policy advocacy role, CeRAI will look to “formulate sector-specific guidelines and recommendations for policymakers”, according to a press statement by IIT Madras.

Balaraman Ravindran, head of CeRAI and Robert Bosch Centre for Data Science and AI at IIT Madras, said upon the inauguration, “It is important for AI models and their predictions to be explainable and interpretable when they are to be deployed in various critical sectors, such as healthcare, manufacturing and banking and finance. They also need to provide performance guarantees appropriate to the applications they are deployed in — which include data integrity, privacy and robustness of decision making.”

Abhishek Singh, managing director and chief executive of the Centre’s Digital India Corporation, said that it is important for policymakers and researchers to “be aware of the risks and challenges while using technologies for solving societal problems, ensuring access to healthcare, making healthcare more affordable, education more inclusive, and agriculture more productive.”

“There is a need for an unbiased and non-discriminatory AI framework as we have unique requirements that require customization as per our requirements,” he added.

To be sure, this is not the first industry-government-academia confluence on the development of responsible AI applications in India. In November last year, policy thinktank Niti Aayog published a discussion paper on the use of responsible AI in developing facial recognition technology infrastructure in the country. The Ministry of Electronics and IT (Meity), along with National e-Governance Division (NeGD) and Nasscom, has also published a responsible AI development ‘toolkit’ to support policy and application development under the ‘IndiaAI’ initiative.

While talks around explainability and responsibility of AI models have led to questions around regulating the nascent technology in Europe, union IT minister Ashwini Vaishnaw said in Parliament on 6 April that the Centre does not plan to adopt legislation to regulate the development of AI. However, he acknowledged ethical concerns around development of AI — which include racial bias, discrimination, violation of privacy and lack of visibility into AI decision-making.

During his response, Vaishnaw added that the Centre is working on standardizing and promoting “best practices” around the development of responsible AI models.