Artificial intelligence has been given a sharper policy push in the Union Budget, with targeted spending across semiconductor manufacturing, data centre infrastructure and AI software development, signalling a more integrated approach to building a domestic AI ecosystem.
On the hardware side, the government has continued backing the India Semiconductor Mission, with the focus shifting to accelerating chip fabrication, advanced packaging and electronics manufacturing.
While no fresh headline allocation was announced for the mission, the emphasis has moved towards attracting investments in packaging, testing and component manufacturing, segments seen as critical for building a viable semiconductor value chain in the near term.
Expanding Manufacturing Schemes
A key announcement was the expansion of the Electronics Components Manufacturing Scheme (ECMS), with the outlay raised to Rs 40,000 crore from about Rs 23,000 crore earlier.
The scheme has already seen higher investments than originally targetted, and the scope has now been widened to cover high-value components used in servers, data centres and AI hardware, including printed circuit boards, power electronics and substrates.
The move is aimed at reducing dependence on imported electronics and deepening the domestic supply chain beyond assembly. The expanded outlay and broader scope could make India more competitive in capital-intensive segments that are critical for AI infrastructure but currently dominated by imports.
Data centres form the second major pillar of the AI push. The government has reiterated policy support for the sector, with tax incentives for greenfield data centre projects. These include a 100% tax holiday for ten consecutive years for eligible data centre investments, along with infrastructure status benefits such as easier access to long-term financing.
The incentives will apply if global firms operate via Indian data centre developers and resellers for hosting global cloud service providers. This is expected to benefit companies building capacity for firms such as Amazon and Google, both of which are expanding their cloud and AI infrastructure footprint in India.
Expanding Compute Capacity
The Budget has also offered customs duty rationalisation for data centre equipment and signals support for reliable power and connectivity, which are seen as key constraints for large-scale AI workloads.
India currently has data centre capacity of about 1,000 MW, and industry estimates suggest this could double over the next three to four years, driven largely by AI-led demand for compute and storage.
The parallel push on electronics manufacturing is expected to lower the cost of servers and related hardware over time, improving the economics of domestic data centre expansion.
The government has indicated that high-performance computing resources will be made available to startups, researchers and academic institutions at subsidised rates to address the high cost of training and deploying AI models. This is seen as critical for enabling domestic AI development beyond large technology firms.
Spending has been stepped up on skilling in semiconductors, electronics manufacturing and AI, with programmes focused on chip design, fabrication processes, data science and applied AI — areas where talent shortages have emerged as a key bottleneck.
In short, the measures point to a shift from policy signalling to ecosystem building. While earlier budgets focused largely on announcing manufacturing incentives, this year’s proposals attempt to link hardware, data infrastructure and software development into a single pipeline.
“Long-term policy certainty recognises that digital infrastructure is now strategic national infrastructure. As AI adoption accelerates across sectors, secure and resilient compute capacity will underpin public services, enterprise innovation, and long-term competitiveness,” Puneet Chandok, president, Microsoft India & South Asia said.