India seeks a seat at the AI high table as leaders converge on Delhi

India positions itself as counterweight to US-China dominance

Next week, global leaders on artificial intelligence (AI) from over 100 countries — including top chief executive officers and heads of state, along with more than 35,000 delegates — will descend on Delhi for the much-anticipated India AI Impact Summit, to be inaugurated by Prime Minister Narendra Modi. 

This is serious business. 

Expectations are that investments adding up to $70–100 billion will be announced during the two-day main summit, spanning the entire technology (tech) ecosystem. Even as the US and China battle for AI supremacy, together capturing 65 per cent of global investment across the AI value chain, India is positioning itself as an alternative force in this global race, seeking a seat at the table where AI’s future agenda will be shaped. 

Explaining the government’s AI strategy, Minister for Electronics and Information Technology Ashwini Vaishnaw said, “Sovereign AI is India’s national goal. Given the size of our country, we need to be competitive across all five layers of the AI stack — applications, models, chipsets, infrastructure, and energy.” 

India faces competition. South Korea, the UK, France, Germany, Saudi Arabia, and the United Arab Emirates (UAE) are all racing to build their own sovereign AI capabilities. Still, several factors tilt in India’s favour. It is among the world’s largest consumers of AI, making it a priority market for global players. ChatGPT lists India as its second-largest market by downloads, after the US. Perplexity, according to reports, has identified India as its fastest-growing and largest market in 2025. For Anthropic, India is the second-largest market for Claude usage, again behind the US. 

US-based ElevenLabs, a leader in AI-generated voice and audio tools, counts India as its largest market by total signups and its second-largest by enterprise revenue. SensorTower data shows India accounted for 16 per cent of the world’s 3.8 billion generative AI app downloads in 2025, placing it at the top globally. 

A Bain & Company analysis released at the World Economic Forum also points to India as a frontrunner in AI adoption. About 86 per cent of firms are exploring or deploying AI, while 14 per cent of the population actively uses AI tools. 

India-focused AI venture capital (VC) funds are now preparing to place larger bets on domestic startups, supplementing government support. Venture Intelligence data shows that in 2025, 16 such funds raised $1.87 billion, up sharply from $358 million the previous year. These included players such as Lightspeed, Nexus, and Accel. 

Challenges remain. India still lacks homegrown AI companies that have achieved meaningful scale — say, revenues exceeding ₹1,000 crore. This raises a fundamental question: will Indian AI startups find buyers for their products, or will enterprises continue to rely on global tech firms? 

There is also concern that India’s information technology (IT) services companies, long dependent on labour arbitrage, have been slow to invest in AI and risk losing ground. The same criticism applies to several large Indian conglomerates. 

Infrastructure is another constraint. India’s data centre capacity — critical for an AI push — accounts for just 3 per cent of global capacity. 

Between 2010 and 2024, India’s cumulative AI investment amounted to only 1.2–1.8 per cent of its 2024 GDP, placing it ninth among 11 countries analysed by Bain, behind Canada, the UK, the UAE, Singapore, South Korea, and, unsurprisingly, China and the US. Although AI-focused VC investment rose to $928 million in 2025 — a 26 per cent increase — it remains insufficient for startups seeking scale. 

Despite this, the government believes it has a workable strategy built around India’s strengths. Vaishnaw argues that 95 per cent of global workloads can be handled by small AI models that are cheaper to build. A model with 50 billion parameters, he says, can meet the needs of enterprises worldwide. The highest returns lie at the top of the AI pyramid — applications and use cases — where India has the wherewithal to lead. 

India already has a large IT services industry with deep knowledge of how enterprises operate globally. These firms are now pivoting towards AI and increased AI hiring by 33 per cent last year. 

That focus lines up with Bain’s projections that global annual investment in AI applications will reach $1.5 trillion by 2030, far outstripping spending on foundational models or infrastructure, while delivering value through domain-specific use cases. In healthcare alone, AI adoption could yield savings of up to 5 per cent. 

Alongside this, the government has supported 12 sovereign AI models with 50–120 billion parameters through grants. These are focused on national priorities such as healthcare, governance, agriculture, and weather forecasting, with multilingual and audio capabilities. Four or five of these, including the much-discussed Sarvam AI large language model, may be showcased at the summit. 

Infrastructure investment is also swelling. Google, Amazon, and Microsoft have announced plans to invest over $70 billion in India over the next few years, largely to build hyperscale data centres. India’s edge lies in cost: it is cheaper to build high-specification data centres in India than in Indonesia, South Korea, Malaysia, Thailand, or Singapore. 

Energy availability adds another advantage. Data-heavy AI servers demand vast amounts of power, preferably renewable. According to S&P, India will have sufficient capacity by 2030, and the additional 15–30 gigawatt hours required by data centres would account for less than 10 per cent of the renewable power expected to come online by the end of the decade. Opening nuclear power to public-private partnerships further reduces energy constraints. 

The recently signed India–US trade deal also helps by removing restrictions imposed by the Joe Biden administration on the number of graphics processing units (GPUs) India could import. With restrictions now relaxed, a major bottleneck in expanding computing capacity has been cleared. 

The government continues to push its design-linked incentive scheme to support chip design firms and has outlined plans to design GPUs domestically — a process expected to take three to four years. Rather than starting from scratch, it aims to build on existing architectures to save time and cost, and has held discussions with Nvidia on collaboration. 

The final — and perhaps most disruptive — question is jobs. Will AI lead to large-scale losses, and how will the government respond? NITI Aayog estimates that India’s tech sector employs 7.5–8 million people, and without proactive intervention, about 1.5 million jobs could be at risk. 

For the government, managing this transition may be the hardest test of India’s AI push.