Artificial intelligence in the country face three structural challenges. This includes dependence on global technology providers, control over sensitive data and the ability to build AI systems that understand the country’s diverse languages and local contexts. KPMG India’s October 2026 report, Sovereign by Design: Building India’s Independent, Inclusive and Future-Ready AI, identifies these pressures as critical to the country’s AI strategy.
“Sovereignty in AI is not about doing everything ourselves. It is about never being in a position where someone else can switch off our ability to serve our citizens,” Nilachal Mishra, Head of Government and Public Sector, KPMG said in the report.
Geopolitics and supply concentration
The first challenge is the concentration of advanced computing infrastructure and cloud services among a handful of countries and companies. The report estimated that the US accounts for about 75% of global AI supercomputer performance, while China accounts for around 15%. It also cites industry research showing that Amazon Web Services, Microsoft and Google together held 63% of the cloud infrastructure market in the first quarter of 2026.
Export controls, licensing changes or service withdrawals could disrupt access to critical technologies. For India, this raises questions about the resilience of AI systems supporting essential public services. KPMG advocates a balanced approach in which India partners with global technology companies while retaining greater control over critical infrastructure and applications.
Data protection and privacy
The second pressure concerns the use of sensitive information, including health records, land records and welfare data, to develop and operate AI systems. India’s Digital Personal Data Protection Act, 2023, and the DPDP Rules, 2025, establish a framework for handling personal data, with most obligations scheduled to take effect in May 2027.
The report argues that governance must be built into AI systems from the beginning. Data residency, audit rights and the ability to move workloads between providers will become important considerations for businesses and government departments.
Making AI work for India’s languages
The third challenge is linguistic inclusion. AI systems may perform unevenly across regional languages, dialects and code-mixed speech, limiting their usefulness beyond digitally fluent, English-speaking users. The report stresses that “AI that does not understand Bharat cannot serve Bharat”. It calls for greater investment in Indian-language datasets, voice-first interfaces and applications that work on low-cost devices and limited connectivity.
For India, AI sovereignty is not simply about developing domestic models. It also requires reliable infrastructure, trusted data governance and technology that can serve the country’s diverse population.
