The biggest AI summit of 2026 just wrapped in New Delhi, and the headline isn’t a new model or a benchmark war. It’s this: 88 countries — including the US, China, and Russia — just signed a shared vision for how AI should be built and governed.

That’s not supposed to happen.

The Declaration Nobody Expected

The India AI Impact Summit 2026 ran February 16–21 at New Delhi’s Bharat Mandapam. Prime Minister Modi inaugurated it alongside Macron and UN Secretary-General Guterres. The guest list read like a tech industry yearbook: Pichai, Altman, Amodei, plus delegations from over 100 countries and 20-plus heads of state.

India’s pitch was simple and sharp: AI shouldn’t be a two-player game between Washington and Beijing. As Jakob Mökander from the Tony Blair Institute put it, “India is right now the player that most confidently says, ‘We reject this dynamic.’”

The resulting New Delhi Declaration — rooted in the Sanskrit principle of Sarvajan Hitaya, Sarvajan Sukhaya (welfare for all, happiness for all) — delivered concrete frameworks alongside the philosophy:

  • Charter for Democratic Diffusion of AI — making foundational AI resources affordable for developing nations
  • Global AI Impact Commons — a platform for countries to share and replicate successful AI use cases
  • Trusted AI Commons — shared tools, benchmarks, and security best practices
  • International Network of AI for Science Institutions — pooling research across borders
  • AI Workforce Development Playbook — preparing citizens for an AI-driven economy

The signatory list is the real story. The US and China at the same table is notable. Add Russia, Iran, Cuba, Fiji, Bhutan, and Suriname, and you’ve got geopolitical breadth that’s almost unheard of in tech governance.

The Money Was Enormous

Investment pledges made the declaration look modest by comparison. Reliance Industries and the Adani Group committed a combined $210 billion in domestic AI and data infrastructure. Adani specifically earmarked $100 billion for renewable energy data centers — a direct response to AI’s growing power consumption problem.

Microsoft pledged $50 billion by decade’s end to bring AI to lower-income countries. Blackstone dropped $600 million. TCS and Infosys announced partnerships with global AI leaders including OpenAI. Total deals announced: north of $300 billion.

For perspective, that figure rivals what the US has committed domestically — and India’s infrastructure costs are dramatically lower. The Global South isn’t asking permission to build AI infrastructure. It’s just building it.

The Chaos Behind the Scenes

Let’s not pretend the summit was flawless. Over 250,000 registered attendees meant overcrowding, hours-long lines, visa nightmares, and traffic that paralyzed chunks of New Delhi.

The no-shows stung. Jensen Huang pulled out. Bill Gates canceled his keynote hours before he was due to speak, officially to “ensure the focus remains on the Summit’s key priorities.” The actual reason: renewed scrutiny over his Epstein ties.

The meme-worthy moment: Altman and Amodei, seated next to each other during a staged unity photo with PM Modi, conspicuously held hands with people on either side — but not with each other.

And in the summit’s most embarrassing incident, an Indian university was reportedly asked to leave after a staff member tried to pass off a Chinese-made Unitree robotic dog as their own creation. India’s opposition Congress party pounced, saying the Modi government had “made a laughing stock of India globally.”

From Safety to Impact — A Telling Rebrand

Here’s what the naming evolution tells you. Bletchley Park 2023: “AI Safety Summit.” Seoul 2024: still safety. Paris 2025: “AI Action Summit.” Now India: “AI Impact Summit.”

That shift from safety to impact reflects who’s steering the conversation. Developing nations care less about hypothetical existential risks and more about how AI can improve healthcare, agriculture, and education right now. The IMF’s Kristalina Georgieva struck the balance, noting AI could lift global growth by 0.8% while warning about job displacement and financial stability risks.

Macron pushed for “let’s do better together” over “let’s do more” — a subtle jab at the US-China arms race mentality that’s dominated AI discourse for years.

Does Any of This Actually Matter?

The uncomfortable truth: the New Delhi Declaration is non-binding. No country is obligated to deliver on anything. We’ve seen this film before with climate agreements.

But dismissing it would be shortsighted. The frameworks — if actually built — could become real infrastructure for global AI cooperation. The Global AI Impact Commons and Trusted AI Commons could be transformative for smaller nations trying to build AI capabilities without starting from zero.

More importantly, this represents a philosophical shift. For the first time, a major AI governance event wasn’t dominated by Western perspectives or filtered through a Silicon Valley lens. India — 1.4 billion people, massive developer community, growing tech sector — is positioning itself as the bridge between the AI haves and have-nots.

Eighty-eight countries agreeing on anything is no small thing. Whether the promises hold is the test that starts now.