Key facts
- BharatGen warns India risks becoming a mere AI consumer without indigenous models
- Foundational AI model development flagged as a strategic national priority
- India's own large-scale AI model efforts remain nascent compared to US and China
BharatGen, the Indian government-supported programme aimed at building homegrown foundational AI models, has sounded a sharp warning: India must invest in developing its own base-layer AI systems or resign itself to being a technology consumer in the most consequential innovation race of the decade.
The caution is pointed and specific. Foundational models — the large, general-purpose AI systems that underpin everything from chatbots to code generators to medical diagnostics tools — are currently dominated by a handful of American firms such as OpenAI, Google DeepMind and Anthropic, as well as a rapidly growing cohort of Chinese companies. India, despite its vast pool of engineering talent and one of the world's largest digital populations, has yet to field a competitive foundational model at global scale.
BharatGen's concern goes beyond commercial competitiveness. A country that does not build its own foundational models will also lack control over how AI systems represent its languages, values, cultural contexts and data. Indian languages, many of which are severely under-represented in Western training datasets, risk being marginalised further if global models continue to be built primarily on English and Mandarin-centric corpora.
The warning arrives as India's AI Policy framework is still taking shape and as global powers accelerate their AI investments. BharatGen's message to Indian stakeholders is clear: the window to build genuine AI sovereignty is open now, but it will not remain so indefinitely.