Respondents reported strong interest in private and sovereign AI, but uneven readiness across security, infrastructure, and governance.

Based on a Sep–Oct 2025 survey of 2,567 respondents (senior decision makers) across 30+ markets* on various aspects of private and sovereign AI (involving control, data locality, security posture, and AI operating models), a global IT services and consulting firm has shared some findings with the media.

First, 95% of respondents had indicated that private or sovereign AI was important to their AI strategy, while 29% had indicated that sovereign AI was being prioritized in a concrete near-term way.

Second, 35% of respondents who were chief AI officers had identified the difficulty of building, integrating, and managing complex AI models in private and sovereign environments as their top barrier to AI adoption, and nearly 60% of AI leaders^ in the data had cited cross-border data restrictions as a major challenge.

Other findings

Third, 38% of respondents had reported high confidence in their cloud security posture, while 48% had “strongly agreed” that they had invested sufficiently in data storage and processing capacity to support AI workloads. Also:

  • 96% of respondents had reported that legacy infrastructure was slowing AI adoption, while more than 99% had reported that they were actively reviewing how to integrate AI into legacy environments.
  • 47% of respondents had reported full confidence in meeting data sovereignty requirements for AI.
  • 51% of respondents had cited integration complexity in hybrid environments as a top challenge when running AI workloads on private cloud, and it ranked as the number-one challenge overall.
  • 97% of respondents had agreed that critical workloads were best kept in private or on-premises environments, with less sensitive or noncore tasks handled elsewhere.
  • 96% of respondents had feared privacy violations and misuse of customer data related to AI and GenAI.
  • AI leaders in the data^ had been 10 to 11 percentage points ahead of other respondents on several prioritization and readiness indicators, including infrastructure and AI strategy alignment.
  • 55% of AI leaders had reported full alignment of IT infrastructure strategy with AI strategy, compared with 45% of all others.
  • 56% of AI leaders had reported centralized AI governance, compared with 38% of all others.

According to Abhijit Dubey, CEO and Chief AI Officer, NTT DATA, the firm that commissioned the survey, organizations with more mature architecture, infrastructure and governance were better positioned to capitalize on enterprise AI.

*Methodology: 30 markets, 12 industries and five regions; a separate respondent base of 211 CIOs was used for the CIO-specific findings. The report also references a broader companion study in the same research series, with nearly 5,000 total decision-makers across the combined studies. The survey report states respondents were from organizations of varying sizes, but the published sample is heavily weighted toward large enterprises, including a substantial share of organizations with more than 10,000 employees. Findings are self-reported and based on prompted survey responses. As with any sponsor-commissioned survey, the framing emphasizes private and sovereign AI risks and the need for architecture, infrastructure and ecosystem redesign, so the results should be read as directional rather than population-wide proof.

^Defined in the full report as participating organizations that had a “well-defined” or “in-progress” AI strategy, a “mature or evolved” AI operating model, and “significantly higher profits” from AI than peers.