India is accelerating efforts to build a sovereign artificial intelligence (AI) technology stack covering compute, data, models, and applications, aiming to cut dependence on foreign platforms and safeguard national security. 

The IndiaAI Mission, led by the Ministry of Electronics and Information Technology (MeitY), has already procured about 34,000 Graphics Processing Units (GPUs) from Nvidia, AMD, Amazon Web Services (AWS), and Intel. Startups like Sarvam, Gnani.ai, and Soket AI are developing homegrown large language models (LLMs).

According to a report by Business Today, experts warn that without domestic capabilities in advanced-node semiconductor fabrication, High Bandwidth Memory (HBM), and chip packaging, India risks losing AI-led economic gains and talent to other nations. 

“Failing to secure the silicon stack means forfeiting Gross Domestic Product (GDP) gains, losing standards-setting influence, and watching our best talent migrate abroad,” said Nitin Mittal, Deloitte’s Global AI Leader.

Dr Vivek Raghavan, co-founder of Sarvam, cautioned against relying solely on foreign AI models, noting that foundational models influence language, decision-making, and worldviews. 

Punit Pandey, founder of Ojas Softech and Astrosage AI, stressed the need for cloud sovereignty to control national data.

Abhishek Singh, Additional Secretary at MeitY and Chief Executive Officer (CEO) of IndiaAI Mission, highlighted Bhashini, a multilingual natural language platform, as an example of AI serving real public needs. 

Experts agree that securing the full silicon value chain is key for India to transform from a global back office into an AI leader.