Nearly all government agencies (94%) are already using generative AI (GenAI) applications and workloads, with the majority of these IT professionals saying the technology supports goals for increased productivity, automation and efficiency, according to the 2025 Global Public Sector Enterprise Cloud Index (ECI) report. And yet, paradoxically, many IT professionals say their organizations lack the internal skills, infrastructure and cybersecurity measures needed to support the technologies implementation and success.
The Nutanix-sponsored ECI report is based on a survey of global IT decision makers, and respondents from the public sector expressed opportunities and challenges they have with embracing technologies such as GenAI and containerization. Less than one-quarter (24%) of respondents say that a lack of use cases is holding back their implementation of GenAI at their agencies. The biggest hurdles are related to data security, privacy, IT skills gaps, and legacy technology.
The ECI report provides recommendations for public sector organizations looking to improve their implementation and success with application containerization and GenAI, including:
- Capitalize on long-term GenAI potential
- Modernize IT infrastructure
- Upskill for AI and cloud workloads
To better understand the ECI findings and recommendations, The Forecast interviewed Nutanix experts who work with public sector organizations, including US Federal government agencies on their modernization initiatives.
Capitalize on Long-Term GenAI Potential
Most respondents (69%) say they expect their organization will see a positive return on GenAI initiatives over the next one to three years. However, the report notes, planning GenAI projects and goals for a timespan of under 12 months may contribute to “underperformance, failure, or abandonment.”
“It’s definitely about long-term outcomes,” said Kanuj Behl, a Nutanix cloud transformation architect working with federal government customers.
“AI projects come with one certainty—they require investment—but their true value only becomes apparent once they're up and running in your organization."
According to ECI findings, 58% percent of respondents expect GenAI to help their agency increase productivity; 50% believe the technology will increase automation and efficiency; 48% stated that GenAI will increase innovation; and 39% agreed that it will reduce costs.
The benefits of GenAI may look different from – and be harder to measure than – those seen in private industry, according to Sherry Walshak, director of public sector solutions marketing at Nutanix.
"Public value isn't about profit margins or stock prices,” she said. “It's about whether government agencies charged with specific responsibilities are actually making life better for people. Are the streets safer? Is the air cleaner? Are citizens getting the help they need without jumping through endless hoops?
“It's also the basics that keep democracy working: being transparent about what the government is doing, following the rules, and earning people's trust every day."
Modernize IT Infrastructure
The report reveals critical infrastructure gaps. For example, containerization is essential for supporting GenAI applications, but 76% of respondents believe their current IT infrastructure needs at least moderate improvement to fully support cloud-native apps and containers. Additionally, 59% cite portability of applications between clouds and on-premises environments as a key challenge.
“There are no legacy AI applications,” noted Tom Bisbee, a Nutanix systems engineering manager for state and local government and education.
“They’re all modern apps, and modern apps run on containers.”
The sheer number of government agencies provides a sense of the scale of modernization required to support cloud-native apps in the public sector. According to the 2022 Census of Governments, there are approximately 90,000 government units across all levels in the U.S. alone, including counties, municipalities, townships and special districts such as school and fire protection districts.
Data governance also presents a significant barrier to innovation, with 67% of respondents citing data silos as a key challenge.
“The biggest challenges I see at the state government level are around siloed data, data ownership, and data governance,” said Stephen Devir, Northeast director of sales for state and local government and education at Nutanix.
“The judicial branch will say, ‘This information is mine, no one else can have it.’ Health and Human Services will say, ‘No, you’re not going to know who we’re giving assistance.’ The challenge is how to break those walls down while safeguarding privacy and complying with regulations.”
Security concerns represent an even greater challenge. Ninety-two percent of public sector respondents agree that their organization could be doing more to secure GenAI models and applications. And 38% identify privacy and security concerns as the top challenge associated with leveraging or expanding GenAI workloads, more than any other barrier.
Upskill for AI and Cloud Workloads
According to the ECI report, IT training is the #1 area that public sector organizations need to invest in to support AI. Only 51% of public sector respondents believe their organization has the necessary skills needed to support GenAI, underscoring the need for strategic upskilling and hiring.
“Agencies are being asked to do a lot more, and there’s typically a huge skills gap,” said Behl.
“It’s definitely a concern, because there are so many different technologies involved in keeping GenAI up and running. It’s like buying a car: you know how to use a car, but you don’t necessarily know how the car works, or all the inner details. However, if you want to troubleshoot the technology, you must have some understanding of it first.”
According to the ECI report, 70% of public sector organizations are already hiring for GenAI skills. This is notable, given that many government agencies hire based on five-year budget cycles, Bisbee said. He added that some current employees will be better positioned than others to quickly get up to speed on the technology.
“Taking somebody who is already developing apps and moving into specifically developing AI apps, is relatively easy,” Bisbee said.
“But taking somebody whose job is just to keep the lights on and maintain your IT environment, and training them how to build and manage AI apps—that’s a different thing.”
Calvin Hennick is a contributing writer. His work appears in BizTech, Engineering Inc., The Boston Globe Magazine and elsewhere. He is also the author of Once More to the Rodeo: A Memoir. Find him on LinkedIn.
© 2026 Nutanix, Inc. All rights reserved. For additional information and important legal disclaimers, please go here.