State governments that are moving toward the use of agentic artificial intelligence (AI) technologies should apply the same oversight and governance lessons developed for generative AI tech, according to a March 3 report from the National Association of State Chief Information Officers (NASCIO).
Agentic AI systems – which can autonomously plan, act, and complete multi-step workflows – offer new efficiencies but also raise risks around accountability, security, and data accuracy, the report says, while emphasizing that many core safeguards related to generative AI remain directly relevant as states move to adopt agentic AI technologies.
Key recommendations in the report include: updating AI policies to address agentic functionality; establishing governance frameworks with privacy and oversight measures; prioritizing data quality; building cross-agency partnerships; rolling out AI incrementally to learn and build trust; maintaining transparency with workers and citizens; and enforcing strong security controls to supervise and stop AI agents when needed.
The report notes that some states are already benefiting from early forms of agentic AI in government services, underscoring the need to review policies and guardrails now.
“Agentic AI is no longer a future concept – it’s happening now,” NASCIO said. “States have an opportunity to ensure that adoption is responsible, safe, and aligned with lessons learned from generative AI.”
By building upon existing frameworks, the report argues, state governments can responsibly leverage agentic AI’s capabilities while minimizing risk and protecting public trust.
“Today, states are heavily involved in artificial intelligence (AI) governance,” the report says.
“We know from NASCIO’s 2025 State CIO Survey that 88 percent have AI responsible use policies, flexible guardrails, security policies and/or ethics requirements,” the report continues, adding, “Given that these statistics are from last year, we can assume these numbers are actually higher today.”
NASCIO’s report on the top 10 state CIO policy and technology priorities for 2026 –published by the group in December 2025 – puts AI tech at the top of that list.
