India's AI safety push: AI governance framework to be released by Sept 28, says Vaishnaw

New Delhi, Sep 18 (PTI) The government will release an Artificial Intelligence governance framework by September 28 that would define safety boundaries to safeguard citizens from AI harm, and outline checks and balances, IT Minister Ashwini Vaishnaw said on Thursday, but added that the framework would not be "prescriptive".

The minister further said India will place strong emphasis on AI safety even at the AI Impact Summit it will host in February 2026, and supports the formation of a common global governance framework for AI, that is consensus-based.

Asked whether countries such as China, which did not participate in earlier summits in the UK, South Korea and France, would be invited to the AI Impact Summit, Vaishnaw said that the ministry is working with the Ministry of External Affairs to decide the list of invitees.

The high-profile summit is expected to draw participation from several nations and heads of state, with further details likely to be announced closer to the event, the minister said.

The AI Impact Summit 2026 to be held in India on February 19-20, 2026, will be inaugurated by Prime Minister Narendra Modi. The 2023 AI Safety Summit took place at Bletchley Park in the United Kingdom, while the 2024 AI Summit was held in Seoul, South Korea. The 2025 AI Action Summit was held in Paris, France.

"As we approach the AI Impact Summit, we would also like to present an AI Governance framework. I would like to thank the Principal Scientific Advisor (PSA) to the Government, Ajay Kumar Sood, to have worked in an extensive manner...over 3000 consultations were held over a long period to come up with an entire framework. So that framework we will be launching in the coming days, hopefully before September 28," Vaishnaw said.

As has been the approach all along, over time, those aspects required for safety of citizens can be converted into law, while the other portions will be under a framework.

India, he asserted, believes in striking a good balance of innovation and regulation.

The AI governance framework that will be released on September 28 will not be "prescriptive", the minister emphasised.

Responding to a query on the AI-generated video depicting Prime Minister Narendra Modi and his late mother, and the measures being taken by the Government to curb such misuse, the minister said that synthetic content and deepfakes created through AI are emerging as a major challenge for society.

"In today's society, it is a big question... on what kind of problem can arise from AI-generated synthetic content and deepfakes, and so the AI governance framework and the upcoming AI Impact Summit 2026, will have a big focus on AI safety, safety of citizens and users," Vaishnaw said.

On what the ensuing AI governance framework would entail, Vaishnaw said it would be aimed at safeguarding citizens from AI harm.

"The boundaries in which AI safety related items should be placed will be clearly defined...If there is a harm coming to a person, how should we deal with it...as we go into development of AI application and more rapid adoption of AI how do we ensure that proper checks and balances are there...that kind of framework," he explained.

He added: "As we have maintained all along over a period of time, whatever is required for safety of our citizens, we will convert those portions into law and other portions we will continue to keep as a framework on which we will continue to work with the industry and with stakeholders". PTI MBI MR