SRINAGAR: The Government of India has outlined a comprehensive roadmap for the IndiaAI Mission, aimed at building a full artificial intelligence ecosystem spanning applications, foundational models, computing infrastructure, semiconductor capability and energy supply. The initiative, launched in March 2024, is designed to create an inclusive AI ecosystem while expanding indigenous technological capacity, the Ministry of Electronics and Information Technology informed the Rajya Sabha.
Responding to parliamentary questions, Minister of State for Electronics and Information Technology Jitin Prasada said the government is developing the AI ecosystem across five key layers — application, AI models, compute, infrastructure and energy — to ensure that artificial intelligence solutions can be deployed at a population scale across sectors such as agriculture, healthcare, education, manufacturing and governance.
At the application layer, the government is supporting the development of 30 India-specific AI applications to accelerate the adoption of the technology in real-world scenarios. Existing AI tools are already being used in areas such as crop advisory systems for farmers and AI-enabled medical diagnostics for early detection of diseases.
In the model layer, India is focusing on building sovereign artificial intelligence models suited to Indian languages and domestic use cases. Twelve organisations and research consortia have been selected in the first phase to develop foundational AI models. Several indigenous models have already been launched, including those developed by Sarvam AI, BharatGen and Gnani. The BharatGen initiative, led by an IIT Bombay consortium, has developed the Param2 AI model that supports 22 Indian languages and demonstrates strong capabilities in coding, reasoning and mathematical tasks.
To address computing requirements for AI training and deployment, the government has made more than 38,000 graphics processing units (GPUs) available to startups, researchers and academic institutions under the IndiaAI Mission. These resources are being offered at an average subsidised cost of Rs 65 per hour — roughly one-third of the global average — to lower barriers to innovation.
Alongside computing capacity, the government is working to strengthen the domestic semiconductor ecosystem under the Semicon India programme. Ten semiconductor projects, including two fabrication plants and eight assembly, testing, marking and packaging facilities, have already been approved with a total investment commitment of about ₹1.6 lakh crore. Commercial production has begun in one unit while pilot production has started in three others. These facilities are expected to produce more than 24 billion chips annually from packaging units and support wafer fabrication capacity of about 55,000 wafer starts per month.
The programme has also supported 105 fabless chip design companies with access to advanced electronic design automation tools and infrastructure. In addition, 315 universities across the country are using chip design tools for training students, with usage exceeding 185 lakh hours. So far, 146 chip designs have been taped out by academic institutions, with 94 successfully fabricated and packaged.
To expand the digital infrastructure required for large-scale AI deployment, the government is encouraging investments in data centres and cloud computing. India’s total data centre capacity has grown from about 375 megawatts in 2020 to around 1,500 megawatts by 2025. The Union Budget for 2026-27 also proposed a tax holiday until 2047 for companies offering global cloud services from data centres located in India in order to position the country as a major global data hub.
Officials said large investment commitments are also emerging across the AI value chain. Announcements made during the IndiaAI Impact Summit 2026 indicated potential investments exceeding $250 billion across infrastructure, hardware, foundational models and AI applications. Venture capital firms have also pledged multi-billion-dollar funding for AI innovation and large-scale training initiatives targeting millions of students and public servants.
The expansion of AI infrastructure is expected to significantly increase energy demand. Government estimates suggest that electricity consumption from data centres could reach 13.56 gigawatts by 2031-32. To address this, the government is planning reliable and sustainable energy solutions for data centres and AI facilities. The recently enacted Sustainable Harnessing and Advancement of Nuclear Energy for Transforming India (SHANTI) Act is expected to support future deployment of small modular and micro nuclear reactors to provide round-the-clock clean power for energy-intensive digital infrastructure.
Authorities are also encouraging the use of advanced cooling technologies to reduce water consumption and improve efficiency in data centres. Techniques such as direct-to-chip liquid cooling, immersion cooling and adiabatic cooling are increasingly being adopted to reduce both power and water requirements.
Alongside infrastructure development, the government said it is focusing on responsible and inclusive use of artificial intelligence. Under the Safe and Trusted AI pillar of the IndiaAI Mission, 13 projects have been selected to address issues such as algorithm auditing, privacy-preserving machine learning, explainability, bias mitigation and deepfake detection.
The government has also issued India AI Governance Guidelines that emphasise human-centric AI development with the principle of “do no harm,” while recommending risk-assessment frameworks and safeguards to ensure responsible deployment of AI technologies.
According to the government, India is also playing an active role in shaping global AI governance frameworks through its participation in the Global Partnership on Artificial Intelligence and through the implementation of UNESCO’s recommendations on ethical artificial intelligence, which stress fairness, non-discrimination and gender equality in AI systems.
Data cited by the government from the Stanford AI Index 2025 shows that India currently leads globally in relative AI skill penetration among women with a score of 1.9, compared with 1.7 for the United States and 1.0 for Canada, indicating stronger representation of women in AI-related skills relative to the overall talent pool.
