OpenAI says Indians under 30 send 80% of ChatGPT messages

OpenAI is getting what every AI lab claims to want: mass adoption that is both measurable and monetizable. According to TechCrunch, OpenAI says Indians aged 18–24 now account for nearly half of messages sent to ChatGPT in the country, with under-30s responsible for 80%. India is OpenAI’s second-largest market, with more than 100 million weekly users, and usage skews unusually toward work: 35% of messages relate to professional tasks, above the global 30%.

The more interesting signal is not “students use chatbots,” but what they use them for. OpenAI says Indians use its coding assistant Codex three times more than the global median, with weekly usage quadrupling since a Mac app launched two weeks ago. Indians also ask roughly three times as many coding questions as the median. That lines up with Anthropic’s separate observation (cited by TechCrunch) that 45.2% of Claude tasks in India map to software-related use cases. India is behaving less like a “consumer AI” market and more like a distributed software factory bolting AI onto everything that can be billed.

This is happening as New Delhi tries to position itself as a principled “third pole” in the US–China technology contest. The New York Times reports that at a major AI summit in New Delhi, Indian officials are pitching the country as a rule-setting intermediary: open to Western capital and models, wary of Chinese platforms, and eager to claim “ethical” leadership.

The practical result looks less like sovereignty and more like compliance as a product layer. OpenAI is expanding its footprint with offices planned in Mumbai and Bengaluru and a partnership with Tata Group to secure 100 megawatts of compute and distribute ChatGPT Enterprise through Tata Consultancy Services, according to TechCrunch. OpenAI also signed deals with firms including Pine Labs, Ixigo, MakeMyTrip and grocery-delivery company Eternal, plus education partnerships aimed at reaching 100,000 students over six years.

The “sovereign” AI stack is increasingly American models, trained elsewhere, run on imported chips and contracted compute, then wrapped in Indian procurement rules, data-localisation requirements, and licensing regimes. The Times notes India’s balancing act between Washington and Beijing; balance, in bureaucratic practice, often means more reporting obligations for everyone who isn’t a ministry.

India’s young users are already moving faster than regulators can count—OpenAI’s chief economist Ronnie Chatterji even framed measurement lag as a policy problem, per TechCrunch. The political temptation is obvious: if the state cannot keep up with adoption, it can at least mandate logs, local storage, audits and “safety” certification. That would be a neat way to turn a generation’s productivity hack into a national paperwork engine.

India’s strongest AI story is not government choreography but individual arbitrage: millions of young workers using foreign models to ship code and documents faster than their employers can rewrite job descriptions. If New Delhi wants to “mediate” between superpowers, it may start by not mediating its own citizens into a compliance queue.