Enterprise Survey Finds 93% are Repatriating AI Workloads or Evaluating a Move Away from Public Cloud

A new independent survey of 203 enterprise IT decision-makers reveals that the era of cloud-first AI strategies is giving way to a more deliberate, workload-driven approach.

According to the Enterprise AI Infrastructure Survey 2026, commissioned by Cloudian Inc., 93% of enterprises have already repatriated some AI workloads from public cloud, are in the process of doing so, or are actively evaluating repatriation. Nearly four in five (79%) have already moved workloads, while 73% plan to further shift toward on-premises or hybrid infrastructure over the next two years.

The survey, conducted in February 2026 via the Centiment research platform, identified three converging forces driving enterprise adoption of on-premises AI: data sovereignty concerns, cloud cost unpredictability, and real-time performance requirements.

Key Findings

• Data sovereignty is the top driver of on-premises adoption. When deploying AI involving sensitive company data, 91% of respondents would choose on-premises, private cloud, or hybrid infrastructure over public cloud. Meanwhile, 74% flagged shadow AI – the unauthorized use of cloud AI tools by employees – as a critical or significant security concern, and 58% reported that data residency concerns have directly delayed or scaled back AI initiatives
• Cloud AI costs are exceeding expectations. Forty percent of enterprises report that actual cloud AI spending exceeds initial projections. Nearly half cite cloud-specific cost unpredictability – including difficulty forecasting total cost of ownership and consumption-based pricing fluctuations – as a barrier to expanding AI adoption
• Performance demands favor on-premises deployment. Three-quarters (75%) of respondents identified current or planned AI workloads that require or would benefit from on-premises infrastructure for acceptable latency performance. Use cases include real-time video analytics, manufacturing quality control, and low-latency transaction processing
• AI budgets are growing to fund this shift. Eighty-six percent of respondents expect their AI budgets to increase in 2026, with 40% projecting increases of 25% or more. Combined with the infrastructure shift toward on-premises and hybrid models, these budget increases signal significant enterprise investment in on-premises AI infrastructure over the coming year

“Enterprises aren’t abandoning the cloud – they’re getting smarter about where AI workloads belong,” said Jon Toor, CMO, Cloudian. “This survey confirms what we’re hearing from customers every day: when sensitive data, predictable costs, and real-time performance matter, on-premises AI infrastructure delivers advantages that public cloud alone cannot match.”

Report Availability: The full Enterprise AI Infrastructure Survey 2026 report, including detailed methodology, charts, and analysis across all findings, is available here.Survey Methodology: The survey was conducted in February 2026 via the Centiment research platform. A total of 203 qualified respondents completed the survey, all screened for decision-making authority over AI strategy, investments, or infrastructure. Respondents represented a cross-section of enterprise industries including technology, manufacturing, retail, financial services, healthcare, and government. Two-thirds (66%) have AI applications running in production.