What is the story about?
Artificial intelligence is fundamentally reshaping how companies and regulators think about data sovereignty, pushing CEOs to look well beyond data localisation laws, Sandip Patel, Managing Director of IBM India, told CNBC-TV18.
Speaking on the back of an IBM report focused on Asia-Pacific enterprises, Patel said AI has changed the economic and governance logic of data, making it critical for leaders to rethink who controls platforms, models and outcomes, not just where data is stored. The report urges CEOs in the region to prioritise sovereign cloud strategies, AI-led workflow transformation and hybrid cloud architectures to unlock the true value of enterprise data.
“For years, digital sovereignty discussions have primarily focused on data residency. AI has fundamentally changed that model,” Patel said. “AI uses data, and data’s value is unleashed where you have solutions and intelligence built on top of it.”
Patel pointed to industry forecasts suggesting that digital sovereignty will soon become mainstream. Gartner estimates that more than 75% of enterprises outside the US will have a digital sovereignty strategy by 2030. According to Patel, this shift reflects the growing need for organisations to demonstrate not only compliance, but also continuous control and accountability in an AI-driven environment.
He said sovereignty today extends beyond the physical location of data and raises deeper questions around governance. These include who operates the platform and under which authority, where AI models run, how inference is governed, who has administrative access, and how compliance is enforced and proven on an ongoing basis.
IBM believes CEOs must now think in terms of three pillars: data sovereignty, technology sovereignty and operational sovereignty. Patel explained that data sovereignty matters because the economic value of data is realised during processing, not storage. Keeping governance under Indian jurisdiction throughout the data lifecycle helps build trust with customers and regulators and ensures benefits are retained locally.
Technology sovereignty, he added, is about choice and flexibility. Technology stacks should be able to move across on-premise, private and hybrid environments without locking enterprises into a single platform. Open-source foundations, transparency and auditability are key to maintaining long-term control.
Operational sovereignty, meanwhile, focuses on resilience. Patel said true operational sovereignty means protecting core processes so that business continuity and uptime are maintained even if there are interventions from external jurisdictions, something global companies have witnessed in recent years.
For Indian enterprises, particularly those in regulated sectors, Patel said the challenge has become more urgent as AI adoption accelerates. “They really need to think about sovereignty beyond where data lives. It now encompasses who controls the models, decisions and outcomes,” he said.
Against this backdrop, IBM plans to make its Sovereign Core offering generally available by the middle of the year. Patel described it as the industry’s first AI-ready sovereign software designed to give enterprises provable control over trusted AI environments. The idea, he said, is to make sovereignty an inherent property of the platform, rather than a contractual assurance.
He outlined key architectural elements such as customer-operated control planes, in-boundary identity and encryption keys, local logs and telemetry, and governed AI inference. These features ensure that AI development and execution remain under local oversight, with full traceability, which is especially important for regulated and essential services.
Also Read | Avoid costly, large LLM buildouts; focus on smaller models: Sridhar Vembu on India AI strategy
Patel said the shift towards sovereign architectures will reshape competition in India and globally, as organisations grapple with overlapping local, cross-border and sector-specific regulations in the AI era. “You need proof of where data resides, who can access it, who administers the platform, where AI models run and how inference is managed,” he said.
On concerns about a potential “SaaS calypse” amid massive AI capital expenditure, Patel struck a measured note. He does not expect a collapse but sees AI as augmenting human work rather than replacing it outright. “AI is more like augmented intelligence,” he said, adding that productivity gains will come through areas such as code development, alongside closer collaboration between people and machines.
Speaking on the back of an IBM report focused on Asia-Pacific enterprises, Patel said AI has changed the economic and governance logic of data, making it critical for leaders to rethink who controls platforms, models and outcomes, not just where data is stored. The report urges CEOs in the region to prioritise sovereign cloud strategies, AI-led workflow transformation and hybrid cloud architectures to unlock the true value of enterprise data.
“For years, digital sovereignty discussions have primarily focused on data residency. AI has fundamentally changed that model,” Patel said. “AI uses data, and data’s value is unleashed where you have solutions and intelligence built on top of it.”
Patel pointed to industry forecasts suggesting that digital sovereignty will soon become mainstream. Gartner estimates that more than 75% of enterprises outside the US will have a digital sovereignty strategy by 2030. According to Patel, this shift reflects the growing need for organisations to demonstrate not only compliance, but also continuous control and accountability in an AI-driven environment.
He said sovereignty today extends beyond the physical location of data and raises deeper questions around governance. These include who operates the platform and under which authority, where AI models run, how inference is governed, who has administrative access, and how compliance is enforced and proven on an ongoing basis.
IBM believes CEOs must now think in terms of three pillars: data sovereignty, technology sovereignty and operational sovereignty. Patel explained that data sovereignty matters because the economic value of data is realised during processing, not storage. Keeping governance under Indian jurisdiction throughout the data lifecycle helps build trust with customers and regulators and ensures benefits are retained locally.
Technology sovereignty, he added, is about choice and flexibility. Technology stacks should be able to move across on-premise, private and hybrid environments without locking enterprises into a single platform. Open-source foundations, transparency and auditability are key to maintaining long-term control.
Operational sovereignty, meanwhile, focuses on resilience. Patel said true operational sovereignty means protecting core processes so that business continuity and uptime are maintained even if there are interventions from external jurisdictions, something global companies have witnessed in recent years.
For Indian enterprises, particularly those in regulated sectors, Patel said the challenge has become more urgent as AI adoption accelerates. “They really need to think about sovereignty beyond where data lives. It now encompasses who controls the models, decisions and outcomes,” he said.
Against this backdrop, IBM plans to make its Sovereign Core offering generally available by the middle of the year. Patel described it as the industry’s first AI-ready sovereign software designed to give enterprises provable control over trusted AI environments. The idea, he said, is to make sovereignty an inherent property of the platform, rather than a contractual assurance.
He outlined key architectural elements such as customer-operated control planes, in-boundary identity and encryption keys, local logs and telemetry, and governed AI inference. These features ensure that AI development and execution remain under local oversight, with full traceability, which is especially important for regulated and essential services.
Also Read | Avoid costly, large LLM buildouts; focus on smaller models: Sridhar Vembu on India AI strategy
Patel said the shift towards sovereign architectures will reshape competition in India and globally, as organisations grapple with overlapping local, cross-border and sector-specific regulations in the AI era. “You need proof of where data resides, who can access it, who administers the platform, where AI models run and how inference is managed,” he said.
On concerns about a potential “SaaS calypse” amid massive AI capital expenditure, Patel struck a measured note. He does not expect a collapse but sees AI as augmenting human work rather than replacing it outright. “AI is more like augmented intelligence,” he said, adding that productivity gains will come through areas such as code development, alongside closer collaboration between people and machines.
