On the morning of February 19, hundreds of delegates stood in long, stalled queues outside Bharat Mandapam, New Delhi’s vast convention complex, baking under a winter sun that felt entirely too warm. There was no food, no water, and no clear word on when the gates would open. Prime Minister Narendra Modi was due to inaugurate the main summit, and security had effectively frozen the venue. Bloomberg reported that delegates, some of whom had flown in from three continents, were simply left waiting.
It was, unintentionally, a perfect metaphor for the India AI Impact Summit 2026: enormous ambition, occasional chaos, and underneath it all, something genuinely significant trying to emerge.
By the time the week ended, that something had. Over six days at Bharat Mandapam — the summit ran a day longer than planned, overwhelmed by public interest — the world’s largest AI gathering produced a multilateral declaration signed by 89 countries, investment pledges exceeding $250 billion, the launch of India’s first homegrown large language models, and a clear message to Washington and Beijing: the future of artificial intelligence belongs to more than two countries.
AI Is Not a Race Between Two Nations
The India AI Impact Summit was the fourth in a series that began at Bletchley Park in 2023, continued through Seoul in 2024, and passed through Paris in early 2025. Each summit reflected the anxieties of its moment. Bletchley worried about existential risk. Seoul focused on safety frameworks. Paris pivoted, somewhat awkwardly, toward commerce.
New Delhi went somewhere different.
As the United States and China battle to dominate artificial intelligence, India used the summit to highlight that there are other pathways to navigate the silicon surge. The theme “Welfare for all, Happiness of all” drew from the Sanskrit philosophical tradition, and it was plastered across banners throughout the capital. Cynics rolled their eyes. But the framing worked, at least rhetorically.
“Long term, it’s good for the world that AI is not just viewed as a race between the U.S. and China, and I think that India is right now the player that most confidently says, ‘We reject this dynamic,'” said Jakob Mökander, director of policy research at one international think tank.
That rejection carried real diplomatic weight. The summit drew over 20 heads of government, representatives from 118 countries, more than 100 global AI CEOs and CXOs, and over 500,000 participants, making it one of the largest AI gatherings in history. French President Emmanuel Macron and UN Secretary-General António Guterres addressed the opening ceremony. The who’s-who of global tech arrived in force: Sundar Pichai, Sam Altman, Dario Amodei, Demis Hassabis, Mukesh Ambani.
It was also critically the first in this global summit series to be hosted by a country from the Global South. That matters more than it sounds.
The $250 Billion Question
The money talked loudly. Reliance Industries pledged $110 billion over seven years for sovereign computing infrastructure, and the Adani Group committed $100 billion by 2035 for renewable energy-powered AI data centres. Google announced an estimated $15 billion investment, including a full-stack AI hub in Visakhapatnam. OpenAI signed a partnership with Tata Consultancy Services, making TCS’s data centre business its first such customer. Anthropic announced a partnership with Infosys and opened an office in Bangalore.
The government set an audacious target: attract $200 billion in AI investment over the next two years. Whether that number is achievable is a separate debate. What is clear is that global technology capital is taking India seriously in a way it simply did not five years ago.
Alphabet CEO Sundar Pichai and OpenAI boss Sam Altman both spoke enthusiastically about India’s advantages, its enormous talent pool and its large consumer market. “The excitement here has been incredible to watch,” Altman told CNBC on the sidelines.
The CEOs also said something more unsettling. Altman claimed the world might be “only a couple of years away from early forms of superintelligence.” Demis Hassabis said artificial general intelligence could be achieved within five years, an apparent halving of his projected timeline from the previous year. Amodei, for his part, suggested that advanced AI could produce 25% annual GDP growth for India. He acknowledged that might “sound absurd.” It did, but no one in the room asked him to leave.
Made in India: The Models Have Arrived
Beyond the deal announcements, the summit’s most consequential development may have been quieter: the emergence of a credible Indian AI technology stack.
Sarvam AI, a Bengaluru-based laboratory, launched a new generation of large language models, including 30-billion and 105-billion parameter models using a mixture-of-experts architecture, as well as text-to-speech, speech-to-text, and vision models. The company also unveiled Kaze, a smart glasses product that Prime Minister Modi tried on at the expo — a rare moment where the theatre of a state summit and the reality of a startup intersected usefully.
The government-backed BharatGen launched Param2, a 17-billion parameter model supporting 22 Indian languages with multimodal capabilities — designed for government services, education, and healthcare. The vision is a sovereign AI stack: trained in India, hosted in India, serving India’s languages and needs.
To power all of this, the government announced plans to add more than 20,000 GPUs to its existing base of 38,000, with a target to cross 100,000 GPUs by the end of 2026. It also unveiled the MANAV framework — Moral, Accountable, National Sovereignty, Accessible, Valid as India’s ethical scaffolding for AI development, and announced IndiaAI Mission 2.0, the next phase of its national AI strategy.
The message was deliberate: India does not want to be merely a consumer of AI built elsewhere. It wants to build.
The Delhi Declaration and Its Discontents
The diplomatic centrepiece of the summit was the New Delhi Declaration on AI Impact, endorsed by 89 countries and international organisations, including the United States and China, making it the broadest multilateral consensus on AI to date. It called for democratic diffusion of AI, equitable access to foundational resources, trusted frameworks, and collaborative research.
It also joined India to the Pax Silica initiative, the U.S.-led coalition to secure semiconductor supply chains — a significant geopolitical signal about which side of the great technology divide India is choosing to stand on, even as it presents itself as a neutral bridge between blocs.
But critics were pointed. Mark Brakel, director of policy at the Future of Life Institute, noted that “so many risks, from child safety to national security risks to loss of control, were discussed in the corridors with greater urgency than ever but didn’t make it to the official outcome.” Others questioned whether the Declaration’s voluntary, non-binding language would ever translate into action. Some policymakers described the event as a natural progression from the Paris summit — one that left them feeling more like a trade fair than a diplomatic gathering.
That tension — between India’s genuine desire to shape AI governance and its equally genuine desire to attract investment and project economic power — ran through the entire week. An analysis by TechPolicy.Press argued that the summit’s structure granted “multinational corporations parity with sovereign governments” through the CEO Roundtable and Leaders’ Plenary, while providing no equivalent high-level platform for civil society, labour leaders, or human rights defenders.
The summit also had its embarrassments. A university exhibited a Chinese-manufactured robot dog as an indigenous creation, sparking viral ridicule on social media. IT Secretary S. Krishnan directed the university to vacate its stall, and an apology followed. A Bengaluru entrepreneur alleged that prototypes were stolen from his booth inside the high-security venue — the devices were eventually recovered by Delhi Police. On the last day, opposition youth activists staged a protest inside the hall. India’s IT minister had already apologised for “problems” on day one.
None of this is entirely surprising for an event of this scale and political symbolism. But it pointed to a gap that India will need to close: between the ambition it projects and the execution it can currently deliver.
What Comes Next: A World Looking Over the Horizon
The baton now passes to Geneva, where Switzerland will host the fifth global AI summit in 2027. By then, the world will have had two years to test whether the Delhi Declaration’s voluntarism produces anything real, whether India’s $250 billion in pledged investment materialises into actual infrastructure, and whether the homegrown models can compete with OpenAI, Google, and the wave of Chinese laboratories advancing rapidly.
The stakes are larger than any single summit. For all the summit’s focus on inclusion and development, some delegates wondered whether the Indian government was only posturing — with little mention made of the possibility that India’s vast technology workforce might be uniquely vulnerable to dislocation by AI tools that the very CEOs onstage were promoting.
That is the honest paradox at the centre of this moment. AI, as Altman and Amodei described it in New Delhi, is moving faster than any governance framework can follow. Superintelligence in two years. AGI in five. GDP growth curves that “sound absurd.” And yet the summit chose as its tagline a phrase from a Sanskrit scripture about the welfare and happiness of all.
There is something worth holding onto in that framing. The Bletchley-Seoul-Paris-Delhi sequence has shifted the conversation from abstract existential risk to practical, measurable impact — healthcare tools in rural India, multilingual models for 22 Indian languages, AI-driven agriculture applications for smallholder farmers. That shift is real progress. The question is whether the world’s governments can move fast enough, and honestly enough, to make that progress mean something before the technology outruns them.
New Delhi made one thing unmistakably clear: the Global South is no longer waiting to be invited to this table. It has built its own.
What it does at that table — over the next twelve months, and the twelve after that — will define not just India’s technological future, but the shape of the world that AI builds.
