India Promises an AI Regulation Consultation Paper Within a Month, With Industry Carrying the 'Bulk of the Responsibility'
IT Minister Ashwini Vaishnaw said on Oct. 8 that India should be able to publish a consultation paper on AI regulation in about a month, focused on safety, a human-first approach, skilling, inclusion and societal harms such as deepfakes, and that industry would carry the bulk of the responsibility. It is a promise of a paper to consult on, not draft rules. MeitY Secretary S Krishnan, speaking at the same World Bank report event, said the time to revisit India's rely-on-existing-laws approach 'maybe' has come. The two gave different GPU procurement figures.
India's approach to AI rules, as its top IT civil servant described it on Thursday, has been to rely on existing laws and voluntary compliance. On October 8 its IT minister put a clock on moving past that, though only as far as a paper to consult on. "In a month's time we should be able to publish the consultation paper," Ashwini Vaishnaw said, according to Storyboard18, adding that it would form the basis for subsequent regulation. Moneycontrol carried the same one-month timeline.
He was speaking at an event for a World Bank report on AI; IndiaAI's account on X described it as the India programme for the World Development Report 2026. "Secretary Krishnan spoke about AI regulation, yes, time has come," Vaishnaw said. "I request the department to come out with a regulation focused consultation paper which focusses on safety, that is the most important point, safety related to AI use."
What the paper is meant to cover
Storyboard18 reports five priorities Vaishnaw named for the framework: safety, a human-first approach, skilling, inclusion, and societal harms caused by AI. "Humanity has to be the most important part of this aspect," he said, per Moneycontrol.
On who carries the load, he was direct: "Bulk of the responsibility will have to be taken by the industry, especially on cyber security, on safety, on user harm, on psychological disorders that might come and on deepfakes" (Storyboard18). "We have always maintained that AI and technologies cannot be regulated just by having a law. We have to bring a techno-legal solution," he said.
None of this is yet a rule. No draft text, obligations, penalties or enforcement body were described, and the timing is the minister's "should be able to."
Krishnan: "Maybe some of those times have come"
MeitY Secretary S Krishnan, speaking at the same event, framed the shift. India had proposed a roadmap to "rely on existing laws and the regulation which is in place, look at voluntary compliances, and then think about regulation at a more advanced stage. Maybe some of those times have come," he said, per Storyboard18, noting that the World Bank report advocates a similar roadmap. "The existential risks now are put out far more explicitly."
Two GPU numbers, two speakers
Both men also spoke about government compute, and their figures are not the same. They may describe different procurements; neither report reconciles them.
| Speaker | What he said (Storyboard18) |
|---|---|
| Vaishnaw | "Our demand estimate is for 10,000 GPUs and the first tender that we are coming out is for 5,000." |
| Krishnan | Requests for proposals went out in the last week "both for a 3,000 GPU capacity, including by SEAC, and a 10,000 GPU capacity, which will be entirely intended for government use"; the 10,000 proposal came out a day earlier. |
Krishnan separately said a term sheet signed during the Prime Minister's visit to the UAE would bring serverless compute to India equivalent to roughly another 10,000 GPUs, and that the government's AI dataset platform had passed 16,000 datasets. Vaishnaw said that for AI adoption across departments, "It won't be a tender process. It will be an empanelment process," and that India could have an indigenous GPU for AI inference in "three to four years time frame."
What a month-out paper commits to
A month-out consultation paper is a modest commitment, but it attaches a date to the shift Krishnan described, from voluntary compliance toward regulation. Placing most of the burden on companies, and naming deepfakes and psychological harm, suggests what the paper may ask of AI providers serving India, from domestic model builders to the foreign labs now selling in-country inference. It comes eight months after India hosted the AI Impact Summit that adopted the New Delhi Declaration.
Limits
- All quotes come from Storyboard18 and Moneycontrol reporting of the event; we did not find a MeitY press release or transcript.
- "Within a month" is the minister's estimate, not a published schedule.
- The GPU figures are each speaker's own and are not reconciled; whether Vaishnaw's 5,000 "first tender" and Krishnan's 3,000 and 10,000 proposals overlap is not clear from the reports.
Sources
- Storyboard18: India to release AI regulation consultation paper within a month: Union Minister Ashwini Vaishnaw (October 8, 2026)
- Moneycontrol: Govt to release consultation paper on AI regulation in a month, says Ashwini Vaishnaw (October 8, 2026)
- Storyboard18: Time to revisit AI regulation may have arrived, says MeitY Secretary S Krishnan (October 8, 2026)
- IndiaAI on X: Vaishnaw keynote at the India programme for the World Development Report 2026 (October 8, 2026)
- Times of AI: Anthropic Sells India on In-Country Claude Inference — Without Its Newest Models
- Times of AI: New Delhi Declaration — Global South Hosts the AI Governance Calendar