India’s AI Mission Built on Five-Layer Strategy, Says Ashwini Vaishnaw

Law Notify : Union Minister Ashwini Vaishnaw on Wednesday said India’s artificial intelligence mission is being developed through a structured five-layer approach that spans the full AI value chain, an approach that is now attracting attention from global industry leaders.

Speaking at an ongoing global summit, Vaishnaw explained that the framework starts with the application layer, focused on solving real-world problems. This is followed by the model layer, where AI systems are built and trained. The third layer covers semiconductors and chips, the fourth consists of digital infrastructure such as data centres, and the fifth layer is energy.

He described energy as a decisive factor in what he termed the fifth industrial revolution, noting that the future expansion of AI will depend heavily on the availability of reliable and affordable power. According to him, India’s coordinated planning across energy, infrastructure and applications has been well received within the global AI ecosystem.

Outlining India’s broader AI strategy, Vaishnaw stressed the importance of public-private collaboration. Addressing a panel on the “Role of AI in Economic Growth and Global Influence” at the World Economic Forum, he said India has created a shared compute facility through a public-private partnership, offering access to nearly 38,000 GPUs.

He said this common compute platform is open to students, researchers and startups at roughly one-third of prevailing global costs. In contrast, he noted, access to high-end GPUs in many countries remains concentrated in the hands of large technology companies.

On regulation, Vaishnaw argued that emerging technologies require a “techno-legal” approach rather than relying solely on standalone laws. He said challenges such as algorithmic bias and deepfakes need technical solutions, including detection tools robust enough to stand up in court. India, he added, is developing technologies to reduce bias, strengthen deepfake detection and ensure proper “unlearning” before AI models are deployed.

The Minister also pointed to a shift in the economics of the fifth industrial revolution. He said future returns will depend more on scalable and cost-effective innovation than on brute-force computing. Citing current trends, Vaishnaw said a significant share of AI development can be achieved using mid-sized models in the 20 to 50 billion parameter range, pushing back against the assumption that progress always requires the most expensive hardware.