Many organizations are deploying private AI, also known as sovereign AI, to maintain control over the infrastructure that powers their models and data. Running AI workloads closer to enterprise data can improve performance, support regulatory requirements, and help manage long-term costs. As infrastructure costs continue to fall, private AI is becoming an increasingly practical complement to cloud-based AI services.
A new study of over 500 senior IT and data professionals finds that 91 percent of enterprises running private AI in production report meaningful use of object storage, the highest overall adoption among storage architectures. 44 percent use object storage extensively and 47 percent use it quite a bit in their AI environments — slightly ahead of file-based storage and well above block-based storage.
The report from cyber-resilient storage software company Scality, with research by Freeform Dynamics, shows 81 percent of enterprises say private AI infrastructure they control is critical to their success, driven by sovereignty, compliance, and data proximity requirements. 57 percent prioritize storage performance to avoid AI bottlenecks, compared with 54 percent citing compute or GPU availability and 52 percent citing network bandwidth, reinforcing that storage can be as critical as compute in production AI environments.
“Most industry discussion frames AI infrastructure as primarily a compute challenge,” says Tony Lock, director of engagement and distinguished analyst at Freeform Dynamics. “This research makes clear that enterprises running private AI in production are dealing with a broader systems reality. Many see a need for simple, scalable architectures that keep data close, support multiple AI genres, and balance performance with governance and cyber resilience across the full pipeline.”
The report also shows 44 percent of enterprises adapt existing compute infrastructure for AI and 42 percent adapt existing storage, while 40 percent purpose-build compute and 39 percent purpose-build storage, showing that tiered, hybrid architectures are the norm rather than greenfield-only deployments.
In addition 40 percent of enterprises cite metadata handling at scale as a bottleneck risk and 38 percent report mixed workload handling challenges, reflecting the need for storage platforms that support both high-throughput training workloads and low-latency inference across the AI lifecycle.
“The data defines the problem, and the platform determines who scales,” says Paul Speciale, chief marketing officer at Scality. “This research validates what we see in the field: production AI success depends on how effectively teams manage and operationalize data throughout the AI lifecycle. Scality provides an S3-native, tiered, cyber-resilient foundation aligned with how enterprises are building sovereign AI today. It delivers the control, predictability, and operational resilience required to scale.”
You can find out more and get the full report on the Scality blog.
Image credit: monsit/depositphotos.com
