NEW DELHI: Six months after the Centre proposed an amendment to regulate artificial intelligence-generated content on social media, policymakers are discussing if such oversight should begin even earlier—by defining AI intermediaries and how they differ from other online platforms.
While the jury is still out on the need for a standalone AI law, lawyers and Big Tech firms have made representations to India’s ministry of electronics and IT (MeitY) and policy whitepapers have been submitted on why AI intermediaries need a separate definition that differs clearly from India’s social media intermediary rules, Mint has learnt.
IT minister Ashwini Vaishnaw said 8 October that his ministry would release a consultation paper for a dedicated . Talk of such a law came in the backdrop of instances of grey-area content action by AI platforms—most notably, Elon Musk’s X-owned Grok and its sexual imagery generation on social media—through the year.
Calls for a European Union-style AI law have split New Delhi. While lawyers say such a law would be in the right direction, Big Tech executives are pushing for “sector-wise AI rules, rather than an overarching law.”
“With AI, the user and the platform share liability of content generated since it is the user that puts in the prompt. Certain court verdicts around the world have underlined that AI platforms do editorialize the content generated to a certain extent, complicating questions around their intermediary status and safe harbour protection,” a top executive at a leading Big Tech company said, requesting anonymity because conversations on the matter are private.
An executive at a second Big Tech firm claimed that while standalone conversations around establishing a new definition for AI intermediaries have so far not taken place, they have been a part of “ongoing discussions with MeitY around content on social media.”
India is one of AI’s biggest markets. OpenAI has said that India has over 100 million weekly active users.
New Delhi relies largely on the Information Technology Act, 2000, and its Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, to determine if a platform is an intermediary. At stake is safe harbour protection, which allows social media platforms to avoid legal and criminal prosecution for content posted and propagated through them.
AI, however, is different. OpenAI’s , Google’s Gemini, Anthropic’s Claude and Meta’s Meta AI use their own data and judgement to generate results, based on prompts by users.
NS Nappinai, senior advocate at the Supreme Court of India and founder of Cyber Saathi, said India needs “a complete overhaul of intermediary laws not just for AI, but for multiple specific use-cases.”
“The ask by AI companies for a special law is justified for more reasons than their stated purpose. India needs laws to enable AI adaptation and protect against misuse. While AI transcends specific verticals and one size certainly cannot fit all, that a forward-looking dynamic law can both enable and protect cannot be gainsaid. While doing so, the legislature needs to look beyond intermediary provisions and formulate tech-agnostic laws,” she said.
New Delhi, however, remains in disagreement that a new definition for AI intermediaries and a new AI law can be separate. A senior official with direct knowledge of MeitY’s stance on the matter, requesting anonymity, said that a key point of conflict is differing viewpoints from Big Tech on how India should regulate AI.
“On the one hand, tech companies do not want an overarching regulation, claiming that nascent technology should not be stifled by law. On the other hand, there are rising calls for giving AI platforms a fresh intermediary definition for safe harbour protection. This requires consensus—two things cannot happen together. MeitY’s stance remains to not be heavily prescriptive in regulating AI, but there is certainly a rising need for regulations,” the official added.
MeitY, Meta, Google and did not respond to Mint’s detailed emails seeking comment on the issue until publishing time. OpenAI declined to comment but pointed to an independent study by policy think-tank The Quantum Hub on the need to define AI intermediaries differently.
According to the study report, the 2026 amendments to the IT rules introduced obligations related to synthetically generated information for intermediaries enabling or facilitating its creation but did not expressly identify which AI actors fall within the intermediary category. However, they route certain AI-related obligations through a legal framework historically organized around services that transmit, receive, store or host information on behalf of another person, according to the report labelled TQH AI Taxonomy, published in September.
“Such an extension of obligations without first identifying which AI actors fall within the law’s scope and whether they possess the functional role and control needed to comply, creates regulatory ambiguity,” it said.
On this note, Nappinai said that a standalone law for AI is desirable as “laws do not need to only be restrictive, but also enabling in nature”—clarifying further that “a sound alternative to this would be to incorporate special provisions in existing laws.”
“MeitY has articulated its support for innovation including for AI tech and a lot was to be covered under its proposed (now discarded) Digital India Act. We have the EU AI Act precedent to build on and go beyond. We can no longer eschew technology but can certainly use laws and regulations to build guardrails,” she said.
