‘AI regulations will ideally be light touch, though harm is concerning’

S. Krishnan, secretary, ministry of electronics and IT. (Photo:  goi_meity/x)
S. Krishnan, secretary, ministry of electronics and IT. (Photo: goi_meity/x)


  • Speaking at CII’s Global Economic Policy Forum, S. Krishnan, secretary, Meity, said that despite recognizing key challenges to be resolved in the evolution of AI in India, the Centre sees AI as a sector where 'the positives outweigh the negatives’

New Delhi: India, in trying to regulate artificial intelligence (AI), will take a light-touch approach to enable innovation, emulating the way the domestic information technology (IT) services industry was allowed to flourish in the 1990s, said S. Krishnan, secretary, ministry of electronics and IT (Meity) on Thursday.

Speaking at the Confederation of Indian Industry (CII)’s Global Economic Policy Forum here, Krishnan said that despite recognizing key challenges to be resolved in the evolution of AI in India, the Centre sees AI as a sector where “the positives outweigh the negatives."

He quoted Gita Gopinath, former managing director of IMF, as saying recently that AI can cause an increase in productivity and economic change similar to what was seen during the industrial revolution.

“This possibility interests us—from India’s perspective, the positives outweigh the negatives in the government’s policy thinking on AI. Clearly, there are safety concerns, with deepfakes and misrepresentations that have already been raised, and will need to be addressed more seriously in the days to come," Krishnan said.

Speaking about the Global Partnership on AI (GPAI) Summit in Delhi on 12 December, Krishnan said that one key goal for India would be to expand the number of members that are already part of the global alliance.

“GPAI is dominated by (developed) OECD countries, but how do you bring more of the global south into this? How do you convert AI and use it for the benefit of larger populations? This will be an important objective of what we’d try to do at the summit," he said.

He added that alongside taking a light-touch approach to regulating AI, the Centre will seek to expand the already existing national programme on AI—“primarily to cover what we can do in compute in conjunction with the world."

Compute, Krishnan noted, is one of the top recognized challenges in the domestic development of AI.

This was addressed on 13 October by union minister of state for IT Rajeev Chandrasekhar, who announced the Centre’s ‘India AI’ strategy. This included a ‘datasets platform’ to organize and provide a centrally collated data pool for Indic languages, and a second platform that will seek to develop indigenous compute power through public-private partnerships.

On 3 November, Chandrasekhar said India highlighted four key concerns and areas of focus as part of being one of 28 signatories of the Bletchley Declaration—the first global cross-border collaborative agreement, held in the UK, to regulate AI jointly.

Industry leaders, over time, have agreed on the need to regulate AI across nations. On 28 August, Brad Smith, president at Microsoft, told Mint that a global regulatory framework on AI could help set guiderails for the global AI industry. Nick Clegg, president of global affairs at Meta, told Mint on 28 September a globally agreed framework on AI development is not only necessary, but “also much desirable."

Underlining key challenges in AI development beyond compute, Krishnan said, “The second challenge is the availability of data—that is something we have in abundance. We’ve also taken positive steps in regulation of data, particularly through the DPDP Act. The third element is skills—here, India ranks fairly high globally. That holds true for India’s overall preparedness rank as well, and even in terms of AI adoption by organizations, we’re one of the best in the world—we already have an advantage in all of this to build on. And fourth is research—we’re working with a number of research institutions both within India and elsewhere. Next week, we’ll have the Global Partnership on AI Summit. India has the chairmanship of GPAI, which currently has 29 countries. We hope to further expand this."

Krishnan also added that foundational AI models indigenously developed with India’s own datasets would be better suited to industrial and national outcomes, in a bid to handle bias. “But eventually, if we are able to train models in that way will be more suited to the rest of the global south as well, that is another important dimension. Our efforts towards Bhashini and a variety of others are in this regard," he said.

Catch all the Business News, Market News, Breaking News Events and Latest News Updates on Live Mint. Download The Mint News App to get Daily Market Updates.


Switch to the Mint app for fast and personalized news - Get App