NEW DELHI: India’s artificial intelligence ambitions are rapidly moving from policy vision to measurable economic promise. But a new report by the standing committee on communications and IT makes it clear that the journey from intent to impact remains uneven.
The report, tabled this week, projects that AI could have contributed as much as $500 billion to India’s GDP by 2025, rising sharply to $967 billion by 2035. That would account for nearly a tenth of the country’s $5 trillion economy target. Alongside this, AI is expected to generate nearly 47 lakh new technology jobs by 2027.
At the heart of this push is the Rs 10,371.92 crore IndiaAI Mission, a multi-pronged strategy spanning compute infrastructure, datasets, innovation, skilling, and responsible AI frameworks. Significant allocations of over Rs 4,500 crore for compute capacity and nearly Rs 2,000 crore for startups signal a clear emphasis on building both foundational infrastructure and a vibrant innovation ecosystem. Backed by more than 900 million internet users and a fast-expanding digital base, India has already emerged as a major AI player. Yet, the report strikes a cautious note on execution.
While flagship initiatives such as IndiaAI, Digital India Bhashini, and AIRAWAT reflect policy momentum, their outcomes are still evolving. For instance, the deployment of over 34,000 GPUs marks progress in compute capacity, but scaling high-end infrastructure to globally competitive levels remains a work in progress. Similarly, the relatively modest number of advanced research fellowships, just over 200, highlights the early stage of India’s deep research pipeline.
Efforts to democratise AI through 27 data labs in Tier-2 and Tier-3 cities indicate a push toward decentralisation, but widespread diffusion is yet to fully take shape. The proposed national datasets platform, envisioned as a backbone for AI development, is still under construction.
Across sectors, AI adoption tells a similar story of promise tempered by patchy implementation. Government departments are experimenting with chatbots and predictive analytics, but integration remains inconsistent.