India’s push for digital sovereignty in artificial intelligence (AI) will require control over computing infrastructure, networks, models, data and talent, rather than merely keeping data within the country, experts said at a technology session organised by KPMG in India at Agrasar Bharat Manch.
The panel on “Digital Sovereignty in AI: Building an Independent and Resilient India” highlighted the need for secure networks, clear governance frameworks and greater domestic capabilities as enterprises and government departments expand their use of AI.
Sovereign AI also depends on ensuring that AI systems operate within defined geographical and regulatory boundaries, said Anoop of Lightstorm, a telecommunications service provider. He said enterprises increasingly wanted assurances that AI prompts and agent actions would follow approved routes and that data would not be transferred outside designated jurisdictions, including during system failures.
Telecommunications networks will play a central role in this transition by providing computing capacity on demand and enabling seamless connectivity between data centres and users, a government representative said during the discussion. The panel also highlighted the importance of keeping end-user data within secure networks.
The discussion comes amid rising investment in India’s data centre infrastructure. Speakers cited more than $90 billion in investments and over 850 megawatts of additional data centre capacity by 2025, while pointing to further capacity planned for AI workloads.
However, infrastructure alone will not ensure sovereignty. Dean, a global government and public sector executive whose surname and organisation were not identified in the transcript, said governments needed to consider three dimensions: data, infrastructure and domestic capabilities. He cautioned against treating sovereignty as protectionism, arguing that the goal should be self-reliance and informed decision-making in a technology ecosystem that crosses borders.
Another speaker, Greg, stressed that AI projects often fail to move beyond pilot stages because organisations do not plan for funding, accountability and adoption at scale.
The panel concluded that India’s approach to sovereign AI must balance security and privacy with flexibility, allowing organisations to choose and change technology providers while retaining control over their data and systems.
Leave a Reply