The push to deploy AI is running ahead of the work needed to secure it. NTT DATA’s global research shows that more than 95% of business leaders see Private AI and Sovereign AI as important to the future of business, yet only about 29% have taken concrete action.
That gap matters because AI is no longer judged only by model performance. Privacy, data sovereignty, and legal compliance across regions are now part of whether AI can be used safely and sustainably at scale.
Security has become the new filter for AI adoption
Many corporate systems were built to move data quickly across clouds and applications. That approach is now under pressure because sensitive data must also follow local jurisdiction rules.
Private AI and Sovereign AI are gaining relevance for that reason. Both approaches are meant to help companies keep innovating while still respecting how data is stored, managed, and transferred.
The survey suggests that most senior leaders understand the importance of data sovereignty. Even so, the pace of actual implementation remains far behind that level of awareness.
Awareness is high, but infrastructure is still lagging
The result is a familiar pattern inside many organizations. Teams are enthusiastic about AI, but the underlying security architecture has not been rebuilt from the ground up.
NTT DATA found that about 35% of Chief AI Officers struggle with the complexity of managing AI models in private environments. At the same time, 60% of AI leaders say they feel pressure from cross-border data restrictions.
Those constraints show that AI adoption is no longer only a technical issue. Companies also have to operate within different rules in every market where they do business.
Confidence in existing defenses is limited as well. Only 38% of respondents said they believe their cloud security is strong enough to handle sensitive data.
Why Private AI and Sovereign AI are not the same
Private AI is focused on protecting internal company data from exposure to outsiders or competitors. Sovereign AI goes further by making sure AI systems and operating environments comply with national or regional regulations.
Abhijit Dubey, CEO and Chief AI Officer of NTT DATA Inc., said forward-looking companies are not treating compliance only as a way to avoid penalties. Instead, they are using this moment to build a stronger business foundation that allows AI to operate flexibly across global markets.
That distinction is important because the pressure on AI projects is shifting. The biggest obstacles are no longer just model quality or performance.
Control, location, and regulation are now part of the AI stack
NTT DATA identified five major shifts that are expected to influence the direction of technology projects. Among them, the main barriers to AI now include computing control and data localization, not only the strength of the model itself.
Network architecture is also being shaped by local law because AI depends on continuous data flow. That makes data storage location and system management inseparable from regional rules.
Companies that redesign their infrastructure earlier are said to commercialize AI at scale faster. Others continue to delay action even though they already understand how important data protection has become.
Partnerships still matter in sovereign AI
NTT DATA also emphasizes that Private AI and Sovereign AI still require a tightly coordinated technology partner ecosystem. In this context, independence does not mean working alone.
Instead, it means building connected systems with strict governance. That approach is intended to let AI move forward without losing control over sensitive data or compliance obligations.
The findings are based on two scientific studies involving nearly 5,000 senior decision-makers. They span dozens of industries across more than 30 global markets, offering a broad view of the pressure now shaping how companies choose between faster AI adoption and a more serious overhaul of data security.
Source: id.mashable.com
