One hundred billion dollars. That’s not a national budget — it’s what a single Indian conglomerate just committed to building AI data centers.

Adani Group dropped the bombshell at the India AI Impact Summit 2026 in New Delhi, pledging to expand its data center capacity to 5 GW of hyperscale, AI-ready facilities by 2035. Liquid cooling, renewable energy, the works. If it happens, it would be the world’s largest integrated data center platform.

“India will not be a mere consumer in the AI age,” said Gautam Adani. “We will be the creators, the builders and the exporters of intelligence.”

Bold words. But here’s the thing — this summit wasn’t just about building. It was also about what AI is about to destroy.

The Numbers in Context

Before anyone gets too breathless about $100 billion, some perspective: Amazon, Google, and Meta are collectively spending $635 billion on AI infrastructure in 2026 alone. AWS, Google, and Microsoft have already promised $67 billion specifically for India.

So Adani’s nine-year commitment is a marathon bet, not a sprint. Whether that’s smart patience or too-little-too-late depends entirely on how fast this race accelerates — and right now, it’s accelerating faster than anyone predicted.

Why Every AI CEO Flew to New Delhi

Sam Altman. Sundar Pichai. Dario Amodei. Demis Hassabis. Mukesh Ambani. Even Emmanuel Macron.

The reason they showed up is simple math. India now has over 100 million weekly active ChatGPT users — second only to the US. Anthropic confirmed India is Claude’s second-largest market. With 1.4 billion people and a median age of 28, India isn’t just a growth market. It’s the growth market.

“Everyone’s coming in because they realize that this is the place to be,” said Lalit Ahuja, CEO of ANSR. No kidding.

India Is Building Its Own AI — Not Just Buying Silicon Valley’s

This is where things get genuinely interesting. India isn’t content to be a customer.

Startups like Sarvam AI and the IIT Bombay-led BharatGen project are building sovereign large language models for Indian languages. The government’s National AI Mission already runs a pool of 38,000 GPUs that companies can rent for ₹65 ($0.72) per hour. At the summit, they announced 20,000 more are coming.

There’s also a fresh $1.1 billion state-backed VC fund for AI startups, and Blackstone just took a majority stake in Indian AI startup Neysa as part of a $600 million raise.

This isn’t nationalism for show. When ChatGPT can’t properly handle Tamil or Marathi, that’s a real market gap. India-specific models trained on local languages and contexts could become the default for the entire developing world.

The Part Nobody Wants to Talk About

Vinod Khosla, founder of Khosla Ventures, stood on the summit stage and said something that made India’s tech sector flinch: IT services and BPOs will “almost completely disappear” within five years.

That’s not a fringe take from an outsider. Khosla is one of the most successful tech investors alive. And he’s talking about gutting the industry that built modern India’s middle class — the Infosys, TCS, and Wipro machine that employs millions.

Indian IT stocks have already been sliding on AI disruption fears. HCL’s CEO poured gasoline on the fire by saying Indian IT companies “will focus on turning profits and not being job creators.”

The irony is brutal: the same summit celebrating AI’s promise is hosting voices warning it could hollow out India’s biggest economic engine. Khosla’s advice to India’s 250 million young people? Stop selling labor. Start selling AI products.

Inspiring and terrifying in equal measure.

The Hindenburg Warning

Meanwhile, Oxford professor Michael Wooldridge warned that the AI industry’s breakneck pace risks a “Hindenburg moment” — a catastrophic failure that shatters public confidence entirely. Easily bypassed guardrails, unpredictable failures, commercial pressure to ship before testing.

This tension — massive investment colliding with mounting anxiety — is the defining story of AI in 2026. India is betting that building sovereign infrastructure early will protect it from the worst outcomes.

What This Actually Means

For the AI industry: India is no longer just a market to sell into. It’s becoming a builder with scale, cheap talent, and serious government backing.

For workers everywhere: If Khosla is even half right, companies that outsource to India need new strategies. Knowledge workers globally should pay attention.

For AI governance: The Global South is demanding a seat at the table. Modi framed AI as “a servant of human dignity, not a threat to livelihood” — a direct challenge to the move-fast-and-break-things playbook.

For investors: The capital flowing into India’s AI infrastructure — $100 billion from Adani, $1.1 billion in government VC, $600 million for Neysa, $15 billion from Google — has reached critical mass.

Bottom Line

India’s play is clear: don’t let the AI revolution happen to you. Build it yourself.

The $100 billion headline is impressive, but the real story is the ecosystem forming around it. Sovereign LLMs, subsidized GPU access, startup funding, and global partnerships are all converging at once.

Can India pull it off? The country has a history of ambitious announcements that stumble on execution. Adani doesn’t even have land secured for these data centers yet. And a 2035 timeline might feel like an eternity in AI years.

But writing off India would be a mistake. A hundred million ChatGPT users, a government building real infrastructure, and the existential pressure of protecting millions of tech workers from disruption — that’s a potent combination.

The AI race just got a lot more crowded.