Federal agencies are under intense pressure to adopt AI, but their data and infrastructure are not keeping pace. This report, commissioned by MinIO via GovNavigators, draws on a mixed-methods survey of 27 federal data and IT leaders and follow-on interviews with senior officials across agencies including the Departments of Agriculture, Commerce, Treasury, NASA, and others. The findings are unambiguous: agencies want to scale AI workloads but lack the foundational capabilities to do so. Infrastructure limitations — particularly around compute and secure storage — are the defining bottlenecks. Agencies are largely stuck in pilot mode, with enterprise-scale AI deployments remaining out of reach. A GAO report corroborates the findings, citing limited access to advanced computing resources and systemic funding gaps. Key themes include: agencies cannot fully calculate their cloud TCO; vendor lock-in and lack of data portability create operational risk; data governance maturity is too low to support reliable AI; and sovereignty requirements are increasingly make-or-break criteria for cloud decisions. The report provides actionable recommendations for both agencies and industry partners.
.svg)
Federal AI readiness is not primarily a technology challenge — it is a data governance and infrastructure modernization challenge, with many agencies still unable to fully inventory their own data assets.
64% of survey respondents cite data sovereignty and residency as top security concerns, driving growing acceptance of on-premises, controlled, and auditable storage environments over public cloud.
Agencies are trapped between cloud cost growth and rising expectations for data custody — hybrid, cloud-agnostic architectures with bidirectional portability are the recommended path forward.
Federal CIOs, CDOs, and agency data leaders, as well as technology vendors and systems integrators developing solutions for government AI modernization and secure data infrastructure.
.svg)
