Between February and April 2026, the IBM Institute for Business Value and Oxford Economics surveyed 1,000 senior executives responsible for AI, data, or technology across 16 countries and 17 industries. The results show that 71% of respondents consider switching their primary AI vendor or model difficult, 68% find it challenging to meet data residency and sovereignty requirements across regions, and 91% say they do not yet fully understand their organization's dependencies on AI vendors, models, and infrastructure.
The risk does not come only from where data is stored. Surveyed organizations experienced an average of six AI-related disruptions over the past two years, and 81% believe a seven-day outage at a primary vendor would have a severe or critical impact. If enterprises check only where their data resides without assessing model replaceability, infrastructure dependencies, service outages, and changes in terms, they may still bear concentration risk in their overseas operations.
Unima recommends breaking an AI sovereignty review into three layers: the data layer records jurisdiction, access rights, and portability; the model layer records vendors, versions, alternatives, and exit conditions; and the infrastructure layer records hosting location, administrative control, redundancy, and switchover testing. Then decide which workloads need greater control based on business criticality, rather than applying the same high-cost specification to every system. This is a practical checklist compiled from IBM's research and does not represent specific client results or legal advice.
Official source: IBM Newsroom, "IBM Study: Limited Control and Rising Dependencies Leave Enterprises Exposed in the Age of AI," 2026-06-17, https://newsroom.ibm.com/2026-06-17-ibm-study-limited-control-and-rising-dependencies-leave-enterprises-exposed-in-the-age-of-ai; the full study, "The Calculus of AI Sovereignty," was published on 2026-06-16. This article was fact-checked on 2026-08-24. The survey results reflect the respondent sample and do not mean all enterprises face the same situation.