Today at the India AI Impact Summit in New Delhi, the focus was moving AI from theoretical potential to large-scale, real-world deployment, with conversations about India’s role as a global AI hub continuing. Below are highlights from some of the biggest stakeholders during the past 48 hours:
$200 Billion in Indian AI Investments
Union IT Minister Ashwini Vaishnaw announced $200 billion in AI-driven investments over the next two years. He said approximately $90 billion of the $200 billion total has already been committed by various companies across “five layers” of India’s stack: infrastructure, energy, compute capacity, models, and end-use applications.
A significant portion is earmarked for data centers and “AI factories,” with the minister saying that about 51% of India’s power generation capacity will be clean energy. The figure also includes venture capitalists’ investments in what he called “deep tech startups” and the scaling of India’s Digital Public Infrastructure (DPI).
Alphabet’s full-stack commitment in India
One of the larger partners is Alphabet, as CEO Sundar Pichai highlighted the company’s $15 billion investment toward foundational AI infrastructure in India, including:
- The America-India Connect Initiative for fiber-optic connectivity via a new subsea cable project that will connect India with the United States, Singapore, South Africa, and Australia.
- Cloud and compute expansion, with new data centers and cloud capacity within India to support the growing demand for AI computing.
- Construction of Google’s largest AI data center hub outside the United States, located in the port city of Visakhapatnam (Vizag).
- Pichai also joined CEO Demis Hassabis of Google’s DeepMind and James Manyika SVP, Research Labs, Technology and Society at Google, to discuss Google’s “full stack commitment to India,” including a $30 Million AI for Science Impact Challenge and $30 Million AI for the Government Innovation Impact Challenge.
- As part of these initiatives, Google DeepMind and the Indian government will work to broaden access to “frontier AI” capabilities for national priorities through a flagship IIT Medas project — a specialized medical AI platform created by the Indian Institute of Technology (IIT) Madras, the Department of Medical Sciences and Technology, and the Centre for Responsible AI (CeRAI).
- Google is also partnering with Atal Tinkering Labs to bring genAI assistance to over 10,000 Indian schools and 11 million students with a focus on robotics and coding.
India to fuel its AI mission with Nvidia
Nvidia announced major partnerships to accelerate India’s AI sovereignty and build local AI factories. Key collaborations include:
- Nvidia deal with Indian cloud provider Yotta Data Services, which will invest $2 billion to deploy an AI supercluster, featuring 20,736 liquid-cooled Nvidia Blackwell Ultra GPUs by August 2026.
- A cloud cluster within Yotta’s infrastructure under a $1 billion commercial agreement; Larsen & Toubro (L&T) will build a gigawatt-scale sovereign AI factory in India, including an expansion of data center capacity to 30 megawatts in Chennai and a new 40-megawatt facility in Mumbai. Part of this deal will include E2E Networks, which will build an advanced Blackwell GPU cluster on its TIR platform to boost local AI compute access.
- For sovereign AI, Nvidia released the Nemotron suite for Indian developers, including the Nemotron-Personas-India dataset—featuring 21 million synthetic personas—to help create population-scale, localized AI models. Infosys, Wipro, Tech Mahindra, and Persistent will integrate Nvidia AI Enterprise software to deploy AI agents for various industries, including finance and healthcare.
- Synopysys, Siemens, and Cadence are using CUDA-X and Omniverse libraries to build software-defined factories for Tata Motors and Havells in India. Nvidia will also partner with the Anusandhan National Research Foundation (ANRF) to offer academic institutions free access to AI software and mentorship.
Governance in the Age of AI
Former U.K. Prime Minister Rishi Sunak discussed AI governance with Meta Superintelligence Lab’s Alexandr Wang during the “Age of AI” fireside chat. Both acknowledged the anxiety that AI infrastructure’s acceleration is causing, and focused on the concept of “trust” and the ways in which attitudes form and shift around AI. Sunak noted that in “Western countries there is more anxiety,” because of a confidence gap, which he said is as much a policy task as a technical one. “How do we get trust up to do untold good. This is more important than the technology itself, as trust governs how successful the tech will be,” said Sunak.
Wang believes that AI can follow the direction of apps like WhatsApp, which he said is “ubiquitous because of its commitment to privacy and building trust.” He noted that the personal agents that will be deployed will know individuals better than even family and friends in terms of health, goals, relationships, and what you want to live a better life. “Because the form we want the technology to take is so intimate and deeply tied to what we think of the future of the consumer life, it’s at the core of the relationship with technology, so developing the technology in partnership with non profits and governing bodies, is the way to garner trust.” The two also discussed safety, and the need for monitoring and testing of models as they achieve capabilities faster than anticipated.
They pointed to the work of the AI Security Institute in the U.K. as an example of how “frontier labs and governments can work together to assure citizens the technology is safe as we enter this phase of dramatic AI progress,” said Sunak, who noted the playbook is emerging: “The right data; compute; appropriate regulation for trust, and public sector use cases, and innovation ecosystem—ingredients for turning ambitions into reality to have impact on people.” He said that playbook will be important in both private and public sectors, which need to build use cases for trust.
Wang and Sunak agreed that every CEO they talk to is thinking about how to deploy AI in their businesses, and how that parallels with the government’s thinking. “All CEOs are thinking about how am I going to be disrupted by AI, disintermediated by AI, how do I use it to get ahead, to be more efficient and to grow? And they are driving projects from the top down—saying this can’t stay with IT dept—has to be driven by the CEO , pick use cases, have accountability, do pilots and scale them up.” Wang said he doesn’t see the transition happening in governments at the same rate or speed as in the private sector. “Just like in an organization, the person at the top has to drive this, said Wang. “We don’t see that transition in governments yet. It has to be driven by the person at the top. You you need to make it clear that the organization will be an AI-first organization. Then you need a bottoms-up bed of adoption and of success stories.”
Great Powers in the Age of AI and Cognitive Systems
During the session, “Great Powers in the Age of AI and Cognitive Systems,” Ketan Patel, Chairman of Force for Good, spoke about findings in a new report, “Technologies Shaping the Future: Civilisational Transition, Strategic Competition, and the New Architecture of Power.” According to Patel, the most important contributor to AI success is “diffusion” to ensure everyone can benefit from AI. He gave a stark warning that currently, AI is not about the systems or the technology, but about “power, enormous wealth, and the ability to control people” in what he calls a huge civilizational shift, as opposed to a technological one. He warned that as AI on our phones, on social media, learns more about us than our family and friends, “it’s important to understand it can go beyond understanding your preferences and start to change your preferences. You become something other than what you were when you started.” That was a stark reminder that the 19 core technologies the new report says will shape our future, can do good if diffused and distributed across the planet. Currently, the U.S. and China are far ahead of the EU. “No one else is even close.”
