What state CIOs need to get right as AI moves to the top of the agenda

Artificial intelligence has claimed the top spot on NASCIO’s 2026 State CIO Top 10 Priorities list, ending cybersecurity’s nearly 12-year run at No. 1. But the change signals less a reshuffling of priorities than a broader transformation: AI is increasingly shaping how states approach everything from cybersecurity and infrastructure to data governance and service delivery.

That shift is also challenging the traditionally deliberate pace of technology adoption in state and local government. As agentic AI advances rapidly, CIOs are under growing pressure to move beyond experimentation and establish the infrastructure, security and governance foundations needed to deploy it responsibly and at scale.

“I don’t necessarily see that AI is replacing cybersecurity as a priority. I see AI changing almost every priority on every CIO’s agenda,” said Kristin Darby, CIO for the State of Tennessee, during a recent panel discussion produced by Scoop News Group for NTT DATA. “CIOs are simultaneously being asked to enable it, govern it, secure it and demonstrate value from it.”

Targeting friction, not flash

Rather than chasing flashy pilot programs, state technology leaders are focusing AI adoption where public friction runs highest, such as navigating benefit programs or wading through complex application paperwork.

“The goal being to put human judgment where it matters, let technology take friction out of everything around it, but keep the person accountable for consequential decisions,” said Clint Dean, Head of Sales for State, Local and Education at NTT DATA.

Building flexible, not fragmented, infrastructure

As workloads shift toward agentic AI systems capable of autonomous action, infrastructure strategy must adapt accordingly. Panelists warned against standing up fragmented tech stacks for one-off AI use cases, urging states instead to invest in shared enterprise capabilities and flexible compute environments.

“The different compute, cost and latency requirements that AI has compared to conventional applications are very different,” said Sezin Palmer, AI Solutions Lead for North America at NTT DATA. “Rather than focusing on a one-size-fits-all solution, building in that flexibility into the strategy for your infrastructure around AI is really important.”

That shifting infrastructure landscape is also redrawing security frameworks, especially around machine identity and authorization. “Identity is no longer just about people,” said Darby. “If an AI agent can take an action on behalf of an employee or interact with another system, we need clear identity, authorization, auditability and accountability.”

Trusted data over perfect data

Legacy data silos also remain a central hurdle for state CIOs, but panelists agreed that leaders shouldn’t stall projects while chasing perfect data cleanliness across the board.

“We don’t need perfect data everywhere. We need trusted, governed data where it matters,” Darby said, urging states to prioritize the data that drives specific business outcomes rather than trying to “boil the ocean.”

Palmer agreed, recommending an incremental approach built around high-value data domains, integration layers and automated tools. “Focusing on those high-value data domains…and getting that data cleaned up becomes something that’s more a part of the process as opposed to a separate step,” she said.

With budgets tight, state leaders are being pushed toward tools that serve multiple purposes. As Dean put it, the decision ultimately comes down to three questions: “Does it reduce risk? Does it produce the outcome that you want to achieve? And then, is it reusable?”

Learn how NTT DATA helps states streamline operations and deliver exceptional constituent experiences.

This video panel discussion was produced by Scoop News Group for StateScoop and underwritten by NTT DATA.