India is racing to build AI independence after China’s DeepSeek exposed its lag in foundational models. Long constrained by underfunded research and development, just 0.65 percent of GDP in 2024 compared with 2.68 percent in China, and a services-first tech culture, India has struggled to invent at scale.
Linguistic diversity adds complexity, with 22 official languages and limited digital data. In response, the government launched the $1.25 billion IndiaAI Mission, deploying more than 18,000 GPUs and backing startups such as Sarvam AI, which is building a 70 billion parameter model for Indian languages.
Over five years, India expects $12 billion in AI and deep-tech investment. While debates over open-source access, compute shortages, and talent migration persist, the push reflects growing ambition and political will.