India's sovereign AI push moves from planning to pilots, but talent and security gaps remain: Report

India's government is moving beyond conversations around sovereign AI and into experimentation, with nearly all public sector organisations surveyed either evaluating the technology, running pilot projects or making investments, according to a study by research firm IDC commissioned by Dell Technologies.

The report, “Building a Sovereign AI Foundation for the India Government”, found that 46% of government organisations are evaluating sovereign AI technologies, while another 46% are conducting proofs of concept. Only 4% have reached the stage of significant investment, suggesting that while momentum is building, large-scale deployment is still some distance away.

The findings come as India ramps up investments under the IndiaAI Mission and seeks to build domestic AI capabilities amid growing concerns around data governance, national security and dependence on foreign technology providers.

As per the report, India’s standout point is its approach to AI sovereignty being built on top of the country's existing digital public infrastructure. The report also  states that India's AI strategy is closely tied to platforms such as Aadhaar, UPI, ONDC and Bhashini, which provide the digital rails for citizen services and data governance. Rather than pursuing technological self-sufficiency, the focus is on ensuring that AI systems, data and digital platforms operate within national governance frameworks while remaining connected to global innovation ecosystems.

Data as a security concern

Data protection remains a key concern. Around 73.3% of government leaders surveyed said sovereign AI is essential for protecting sensitive national data and complying with local regulations, while 70% said investments in sovereign AI could improve resilience against geopolitical risks and supply-chain disruptions.

National security and defence emerged as the most important use case, with 46% of respondents identifying it as the area where sovereign AI could deliver the greatest value. Procurement and supply chain monitoring followed at 36%, while IT and digital services ranked third at 32%.

The survey also found strong optimism around the next wave of AI technologies. Nearly 98% of respondents said agentic AI could help accelerate adoption. Of these, 44.4% believe it will play a major role, while 53.3% said its adoption should be accompanied by strong oversight and governance safeguards.

Yet the report suggests that the biggest obstacles to scaling sovereign AI may have less to do with technology and more to do with execution. More than nine in ten government leaders highlighted shortages of specialised digital talent as a concern. Network management and integration specialists emerged as the hardest roles to fill, followed by AI safety researchers, AIOps professionals and experts in sovereign data governance.

Cybersecurity is another challenge. Security vulnerabilities in AI systems were cited by 36% of respondents as the biggest barrier to adoption, while 34% pointed to the complexity of complying with different national and international regulations.