New Report Highlights AI Sovereignty Challenges

NTT DATA, a global provider of  AI, digital business and technology services, has released new research showing that enterprise AI is rapidly outgrowing the architecture and infrastructure supporting it, even as data privacy and sovereignty requirements become increasingly stringent. The findings reveal a widening divide between organisations redesigning their AI systems around control, locality and security, and those continuing to layer AI into environments never built to support such demands.

For years, enterprise architecture was designed to move data seamlessly across systems, applications, clouds and borders with speed and efficiency. However, the rise of AI is exposing the limitations of that model. Sensitive data now requires stronger protection, workloads must operate within clearly defined jurisdictions, and AI models must function under tighter governance controls. Data can no longer move as freely as many AI systems require, turning jurisdiction into a critical architectural consideration and elevating the importance of private and sovereign AI.

According to NTT DATA’s 2026 Global AI Report: A Playbook for Private and Sovereign AI, there remains a significant gap between what organizations recognize as necessary and what they are prepared to implement. More than 95 percent of respondents acknowledged the importance of private and sovereign AI, yet only 29 percent said they were prioritizing sovereign AI in a concrete and near-term manner. About 35 percent of Chief AI Officers identified the challenge of building, integrating and managing complex AI models within private or sovereign environments as the leading barrier to adoption, while nearly 60 percent of AI leaders cited cross-border data restrictions as a major challenge. In addition, only 38 percent expressed high confidence in their cloud security posture, despite its importance as the foundation for both private and sovereign AI.

The report distinguishes between private and sovereign AI, noting that while they are closely related, they serve different purposes. Private AI is primarily focused on protecting sensitive enterprise data, restricting access and minimizing exposure, whereas sovereign AI is designed to ensure that AI systems, data and operating environments comply with jurisdictional, regulatory, national or regional control requirements.

“As AI evolves, private and sovereign approaches are testing enterprise readiness,” said Abhijit Dubey. “The organizations that are succeeding are going beyond regulatory compliance and risk mitigation. They are building the operating foundation for AI that can perform across markets, jurisdictions and business environments. Our research shows AI leaders are pulling ahead by treating architecture, infrastructure and governance as strategic requirements.”

The report further identifies five major shifts shaping the next phase of enterprise AI. First, AI is encountering a critical barrier that is no longer related to model performance alone, but rather to the growing need for greater control over compute, data access, security and locality. This is exposing the weaknesses of infrastructure originally designed for centralized and borderless data flows. Second, data jurisdiction has emerged as a core architectural constraint, influencing where data resides, where models operate and how AI systems are governed.

Third, while nearly all organizations recognize the significance of private and sovereign AI, only a minority are taking decisive action toward implementation. Fourth, leading organizations are redesigning infrastructure, governance and operating models early, enabling them to scale AI deployments more effectively and creating a widening competitive advantage over slower-moving peers. Finally, although private and sovereign AI are often associated with independence and control, in practice they rely heavily on tightly coordinated ecosystems, with more than half of organizations citing integration complexity as their greatest challenge.

The findings suggest that private and sovereign AI are fundamentally reshaping how AI systems are developed, governed and scaled. Organizations that redesign their architectures early are positioning themselves to succeed in regulated, distributed and data-sensitive environments, while those attempting to retrofit AI into legacy environments may struggle to generate long-term value from their AI investments.

The report is based on two studies involving nearly 5,000 senior decision-makers across more than a dozen industries, over 30 markets and five regions, and forms part of NTT DATA’s broader global research series examining the strategies separating AI leaders from the rest of the market.