India To Bring New AI Law? What Top Government Adviser Told NDTV

Ajay Kumar Sood, the Centre's Principal Scientific Adviser, said the government is open to bringing in a separate law for artificial intelligence (AI). His statement comes as more people become concerned about the harm AI can cause.

In an exclusive interview with NDTV, Sood also said the Centre has started preparing a paper to ask people for their views on rules for AI.

He stressed that AI offers many new opportunities, but its growing risks mean that carefully planned safety measures and rules are needed.

“Over the past one or two years, things have moved much further than before. The risks are becoming clearer, and we need to do something. So, the idea is to create specific rules and safeguards,” he said.

Researchers and safety experts at leading AI companies, including OpenAI and Anthropic, have recently warned strongly about the possible dangers of increasingly powerful advanced AI models.

India has been making its IT rules stricter to control AI-generated content. In February, the Centre introduced tougher requirements for online platforms to deal with AI-generated and synthetic content, including deepfakes. The government said platforms such as X and Instagram must remove such content within three hours after it is flagged by an authorised government body or a court. The government also announced changes to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, which officially define AI-generated and synthetic content.

“From the very beginning, we have been working on how AI can benefit our society while also avoiding the clear risks,” Sood told NDTV.

He also said that AI agents have created a “new challenge”. He was referring to two OpenAI models that escaped from a closed testing environment and, on their own, broke into the internal systems of Hugging Face, a website where AI developers store and share code.

“If these AI agents are left to act on their own, they can cause more harm,” Sood said.

He said the Indian Computer Emergency Response Team (CERT-In), the government's cybersecurity agency, had also issued advice on using AI agents safely. It said a human must remain involved and approve important actions.

Sood also said CERT-In had advised that every AI agent should be kept separate in a sandbox environment, a restricted space that helps limit what it can access or do.