India may see an investment of up to $150 billion in AI infrastructure by the end of 2026, Ashwini Vaishnaw, union minister for information technology, railways and information and broadcasting, said in an exclusive interview with CNBC TV18 at the World Economic Forum in Davos, Switzerland.
Vaishnaw said India has already secured investment commitments worth $70 billion, with a further $50–80 billion likely over the next 12 months.
The past year has seen a wave of major investment announcements. Google committed $15 billion to set up an AI hub in Visakhapatnam, Microsoft raised its India investment commitment to over $20 billion through 2030, and Amazon pledged $35 billion in investments over the same period.
Vaishnaw also confirmed that India’s homegrown large language models (LLMs) will be unveiled at the upcoming India AI Summit. IndiaAI Mission CEO Abhishek Singh had earlier told AIM that at least two LLMs—developed by Sarvam AI and BharatGen—would be launched ahead of the summit.
The minister said India will eventually develop 12 foundational AI models, each ranging between 50–120 billion parameters, designed to run on relatively small GPU clusters. This, he said, would enable low-cost AI services at scale. Vaishnaw noted that early live testing of Indian models has shown “very encouraging” real-world performance.
Under the IndiaAI Mission, the government has selected 12 startups and organisations to build sovereign foundational models: Sarvam AI, Soket AI Labs, Gnani.ai, Gan.AI, Avataar AI, BharatGen, Fractal Analytics, Tech Mahindra (Maker’s Lab), ZenteiQ Aitech Innovations, Genloop Intelligence, NeuroDX (IntelliHealth), and Shodh AI.
Addressing industry leaders, Vaishnaw urged companies to play a more active role in skilling India’s youth, calling for collaboration on curriculum development to create a robust AI talent pipeline.
