Artificial intelligence is increasingly shaping national security, economic competitiveness and technological self-reliance. As countries invest in sovereign AI capabilities, the focus has expanded beyond developing advanced models to building the institutional, technological and governance frameworks needed to deploy AI securely and at scale.
A report jointly developed by IITM PRAVARTAK – the Technology Innovation Hub of IIT Madras – and Securin Labs examines the policy, infrastructure and institutional measures that could support India's sovereign AI ambitions. The report outlines recommendations spanning Centres of Excellence, AI literacy, public-private collaboration, cybersecurity and large-scale AI adoption across sectors.
In this conversation, Lt General Karanbir Singh Brar, PVSM, AVSM (Retd.), Distinguished Strategic Advisor, IIT Madras PRAVARTAK, discusses the report's key recommendations and the measures he believes are necessary for India to build a resilient and sovereign AI ecosystem.
1. Your report recommends creating a network of specialised Centres of Excellence for sovereign AI. Why is institutional capability just as important as technological capability in India's AI journey?
Technology alone has never made a nation sovereign. Institutions do. India's COVID response demonstrated this clearly. Covaxin, CoWIN and the world's largest vaccination programme were possible not because we suddenly acquired new technology, but because institutions such as ICMR, the Serum Institute, UIDAI and our digital public infrastructure had been built over decades. AI sovereignty requires exactly the same approach.
Centres of Excellence must therefore be far more than research laboratories. They should become national mission institutions with technical depth, legal authority and sustained funding. Their role should include developing sovereign AI models, certifying AI systems before deployment, conducting adversarial testing, setting evaluation standards and bridging research, industry and government. Institutional capability ensures continuity beyond political cycles and creates national competence that compounds over decades.
2. Building sovereign AI requires significant investment in compute infrastructure, talent and research. How should India balance public investment with private sector participation to accelerate this agenda?
Sovereign AI cannot be built entirely by the government, nor can it be left entirely to market forces. It requires a joint approach & a national partnership where the government builds strategic public goods while industry drives innovation and scale. Government should invest in sovereign compute infrastructure, semiconductor capability, national datasets, AI security, Centres of Excellence and long-term research. Industry, on the other hand, should lead product innovation, application development, commercial deployment and global market expansion.
India's digital success stories provide the blueprint. Aadhaar, UPI and CoWIN were public digital platforms that unlocked enormous private innovation. AI should follow the same model.
If India succeeds in combining public purpose with private entrepreneurship, we can create perhaps the world's most vibrant sovereign AI ecosystem.
3. The report highlights AI literacy as a national security priority rather than just an education initiative. Why is improving AI awareness among policymakers, enterprises and citizens so critical to India's long-term AI ambitions?
AI is going to impact every aspect of life and has ramifications at the global scale, impacting national security, economics, development and growth. Hence, it is not just another education initiative but a national security priority. Our work on this report with Securin Labs highlighted that the challenge is no longer access to AI technologies, but the ability of leaders and institutions to understand, evaluate and deploy them responsibly. Today many leaders understand that AI is important but lack the knowledge to distinguish genuine capability from marketing hype. This creates strategic vulnerability. Organisations delay adoption, make poor investment decisions or become dependent on foreign technologies because they cannot independently evaluate them.
That is why I describe AI literacy as national preparedness. Policymakers need decision-grade AI understanding. Business leaders need strategic AI literacy. Engineers require implementation skills. Citizens must understand how AI affects their daily lives and how to recognise misinformation.
During COVID, misinformation slowed vaccination efforts despite scientific success. In the AI era, misinformation could delay national transformation itself. AI literacy therefore becomes as important if not more as digital literacy was two decades ago.
4. One of the more interesting ideas in your report is the '1001 AI Use Cases' framework. Why do you believe widespread AI adoption across MSMEs and public services is more important than focusing solely on frontier AI models?
Frontier models are essential, but frontier capability alone does not transform economies. Diffusion does. History shows that countries benefit most not necessarily by inventing technology first, but by adopting it fastest and widest. India's greatest digital successes, Aadhaar, UPI and CoWIN, were not merely technological achievements; they became national platforms because they reached hundreds of millions of people.
The '1001 AI Use Cases' framework is designed to democratise AI adoption. Every MSME, district administration, hospital, farm, school and public service should identify practical AI applications that improve productivity. Thousands of small successes create confidence, generate uniquely Indian datasets, build local talent and stimulate innovation.
A frontier model sitting in a laboratory has limited impact. But millions of businesses using AI to improve manufacturing, logistics, agriculture, healthcare and governance can fundamentally reshape India's productivity. Sovereign AI ultimately succeeds when it improves the lives of ordinary citizens, not merely when it produces benchmark-leading models.
5. You argue that India's sovereign AI journey could become a model for other developing economies. How can India position itself as a trusted global exporter of AI innovation, cybersecurity expertise and human-centric AI solutions?
India's greatest comparative advantage is that it is solving problems that much of the Global South also faces, large populations, linguistic diversity, resource constraints and the need for inclusive development. If India successfully builds AI for multilingual education, agriculture, healthcare, governance and financial inclusion at national scale, those solutions will naturally be relevant across Africa, Southeast Asia and Latin America. That creates a powerful export opportunity based on trust and shared development challenges rather than technological dominance.
The second opportunity lies in cybersecurity. As AI-driven cyber threats increase globally, India can become a trusted provider of adversarial testing, AI assurance, vulnerability intelligence and secure AI governance. Countries will increasingly seek partners that can help them deploy AI safely, not merely build larger models. India's strategic differentiator should therefore be human-centric, secure and trustworthy AI. That combination aligns with our democratic values, digital public infrastructure and longstanding reputation as a technology partner to the developing world.
6. If you were advising the Government of India today, what are the three immediate actions that would have the greatest impact in putting the country on the path to becoming a sovereign AI leader by 2030?
The first priority is to establish sovereign AI infrastructure. That means rapidly expanding national compute capacity, strengthening semiconductor capability, creating secure sovereign cloud infrastructure and establishing a nationwide network of AI Centres of Excellence with long-term funding and statutory authority. Without infrastructure, sovereignty remains aspirational. Second, launch a nationwide AI literacy and workforce mission. Every policymaker, civil servant, military leader, engineer, entrepreneur and student must become AI-ready. Human capital will determine whether India merely consumes AI or leads it.
Third, accelerate nationwide AI diffusion. Launch the '1001 AI Use Cases' programme, create AI adoption incentives for MSMEs, modernise public services through AI and ensure every sector—from agriculture and healthcare to defence and manufacturing, begins deploying AI at scale. The greatest strategic advantage will come not from having a few exceptional AI systems, but from making AI a national capability available to every citizen and every enterprise.
If these three actions are pursued concurrently, as India did during the COVID response, we can realistically position India not merely as an AI user, but as one of the world's leading sovereign AI powers by 2030.
