Satya Nadella announces Microsoft's $3 bn investment in India after meeting PM Modi

Nadella said, "I am really really excited to announce the single largest expansion we have ever done in India by putting $3 billion additional dollars to expand our Azure capacity."

He said, "India is rapidly becoming a leader in AI innovation, unlocking new opportunities across the country.

"The investments in infrastructure and skilling we are announcing today reaffirm our commitment to making India AI-first, and will help ensure people and organizations across the country benefit broadly."

Nadella made the announcement days after meeting prime minister Narendra Modi in New Delhi.

After meeting the Indian PM, Nadella had posted on X, "Thank you, PM @narendramodi-ji for your leadership. Excited to build on our commitment to making India AI-first and work together on our continued expansion in the country to ensure every Indian benefits from this AI platform shift."

Microsoft will also support the country’s long-term competitiveness by training 10 million people over the next five years with AI skills, as part of the second edition of its ADVANTA(I)GE India program.

Microsoft will expand its cloud and AI infrastructure across datacenter campuses in the country.

Microsoft already has three datacenter regions in the market, and the fourth ready to go live in 2026.

This investment aims to develop a scalable AI computing ecosystem to meet the growing demands of India’s rapidly expanding AI start-ups and research community.

Microsoft Research (MSR) Lab has also announced an AI Innovation Network to deepen its commitment to cultivate the AI ecosystem in India.

Under the AI Innovation Network, MSR will build new collaborations, especially with digital natives, to accelerate the transition from research to real, usable business solutions.

MSR India has already initiated a collaboration with Physics Wallah on math reasoning and is in ongoing discussions with other Digital Natives on topics such as causal inference, optimizing Indic LLMs, prompt optimization, and reinforcement learning.