The State AI Officer Mandate Without the Machine: What NASCIO's First CAIO Survey Reveals
Back to Signal
State & LocalAIGovernmentGovtechModernization

The State AI Officer Mandate Without the Machine: What NASCIO's First CAIO Survey Reveals

November 18, 2026Jess Loban

From Title to Role: Three Years of Rapid Expansion

When NASCIO began tracking state AI leadership in 2023, fewer than 20 percent of states had a state chief AI officer or a comparable title. By 2026, at least 46 states have identified someone tasked with directing artificial intelligence, and 25 have an official with "AI" in their title, according to the survey. That growth tracks closely with AI ranking number one on the NASCIO Top Ten for state CIOs for the first time — displacing cybersecurity, which had held the top spot for twelve consecutive years.

The expansion moved quickly in part because state CIOs are generally not waiting for legislation or executive orders — 79 percent of CAIO positions were created by CIO directive, with 13 percent each established through executive order and legislation (respondents could select multiple). That creation mechanism matters structurally. A role established by directive lives at the discretion of the CIO who created it; it carries no inherent budget authority, and its continuity depends on the next CIO choosing to preserve it. Seventy-six percent of CAIO survey respondents said they were the first person to hold their role as currently defined in their state; most have no predecessor to inherit context from and no institutional record for what the role is supposed to do.

What "Established" Actually Means

The survey asked CAIOs to characterize the maturity of their enterprise AI program. The distribution is instructive: 56 percent said programs are "established but still maturing," 28 percent said they are "in the process of being developed," and only 9 percent described their program as "established and operational." Six percent remain in an early exploratory stage.

That sits alongside a finding from NASCIO's companion 2026 State CIO Survey that 98 percent of states have enterprise AI policies and procedures — a number that sounds like comprehensive readiness. The two data points describe different things. Having a policy is not the same as having an operational program. One state CIO quoted in the CIO survey offered the clearest version of that gap: "Virtually all options were selected with the caveat that these efforts are very early, immature and/or not strictly enforced. We still lack broad understanding and socialization of our AI policy, making enforcement and governance difficult."

A policy that cannot be enforced is a commitment, not a control.

The Governance Architecture Decision

Fifty-nine percent of CAIO respondents described their state's AI governance as hybrid or federated; 31 percent said it was centralized; 9 percent called it decentralized, according to the NASCIO AI Officer Survey.

This is a live operational design decision with direct consequences for what a CAIO can accomplish. A centralized model gives the CAIO authority to set and enforce enterprise standards across agencies. A federated model distributes AI ownership to individual agencies, with the central office providing frameworks and review processes but not control. The hybrid model attempts both, which is the most common approach and also the most capacity-intensive to operate: it requires a central team that can coordinate and influence agency AI programs it does not directly control.

The report argues directly: "[C]hief AI officers would benefit from having full authority to establish and enforce enterprise AI standards. This would reduce agency and staff burden, increase interagency cooperation and unify agencies in the state." That is a case for strengthening central authority — and the structural picture is uneven. The survey found that 72 percent of CAIOs have authority over all executive branch agencies, 44 percent can independently set enterprise AI policy (with another 16 percent sharing that authority), and 53 percent have formal enforcement authority. The other half operate in an advisory role.

That gap matters in practice. CAIOs without enforcement authority are expected to produce consistent, coordinated outcomes across agencies without the structural tools to compel them. That is a governance design problem, not a talent problem.

Three Structural Deficits

The survey makes the capacity gaps explicit across three dimensions.

Staff. Seventy-five percent of CAIOs cited lack of staff as a challenge — the most commonly reported one. Seventy-eight percent named adequate staffing as their top need. State AI governance at this stage requires people who can review agency use cases, manage vendor relationships, coordinate procurement language, run incident response for AI failures, and monitor production systems — simultaneously. A CAIO without a team is writing policy documents, not operating a program.

Budget. Fifty-six percent of CAIO respondents operate with no dedicated AI budget, meaning program costs are likely absorbed in existing agency allocations or funded through one-time sources. A program without a stable budget cannot staff consistently, cannot retain contracted expertise across fiscal years, and cannot make the infrastructure investments that moving from pilot to production requires. The companion CIO Survey reinforces the gap: only 28 percent of states report dedicated AI funding, even as 82 percent report GenAI projects in production — a structural tension between operational commitment and resource allocation.

Authority. Most CAIO positions were established by CIO directive rather than by statute or executive order. That creation mechanism determines what the CAIO can actually require of agencies. A directive-based role can produce guidance; it cannot compel an agency to adopt a governance framework, complete a use-case inventory, or submit AI deployments for review unless the state's IT governance structure already provides that authority. In federated models where agencies have independent budget authority, a CAIO without formal cross-agency enforcement powers is in an advisory position regardless of their title.

The Transition Variable

Thirty-six gubernatorial elections took place in November 2026, according to the NASCIO 2026 State CIO Survey. The incoming class of governors represents a significant transfer of state executive authority, and AI governance programs that exist as CIO-directed initiatives are, by default, in line for review or discontinuation when a new administration takes office.

The AI Officer Survey captures this risk indirectly. With 76 percent of CAIOs as the first to hold their role as currently defined, there are no established precedents for how the role transfers. Without a formal charter, there is no documented scope, no standing governance committee, and no enterprise AI strategy tied to anything more durable than the current administration's priorities.

What survives transitions is what has been institutionalized: programs established by statute, enterprise vendor contracts with governance performance requirements, standing cross-agency bodies with defined membership, and AI-use inventories maintained in enterprise systems rather than a CAIO's personal workspace. Programs that exist as documents on a CIO's drive tend to get replaced by the next CIO's documents.

Incoming state IT and technology advisors to governor-elect offices have a narrow window before the new CIO is in seat to assess what AI governance infrastructure actually exists and what requires explicit commitment to preserve. The NASCIO survey creates a useful baseline for that assessment.

Decisions Before the Transition

The survey data translates to three structural decisions state executives and incoming administrations should resolve before January 2027:

  1. Is the CAIO role established by legislation, executive order, or CIO directive? If CIO directive only, the program's durability depends entirely on who is appointed as the next CIO. A formal charter or executive order creates a foundation independent of that appointment.
  1. Does the governance model match the CAIO's actual authority? A federated model assigns AI ownership to agencies that the central office cannot compel. If the CAIO cannot require use-case disclosure, review, or remediation, the state has a policy framework rather than enterprise governance. Aligning the model with the authority often means either granting real enforcement power through explicit CIO policy or consolidating governance toward a more centralized model.
  1. Is there a defined AI budget, or is funding approval-dependent? A CAIO without a budget cannot retain staff, sustain contracts, or make consistent infrastructure investments. Even a modest dedicated allocation is structurally different from project-by-project approvals: it creates a resource baseline the program can plan against and that an incoming administration can evaluate rather than simply inherit as an unfunded obligation.

None of these are technology decisions. They are organizational decisions that determine whether a technology program operates or simply persists.

Sources and further reading

Share this article
LinkedIn

BUILD WITH US

Ready to Solve Hard Problems?

Spartan X builds AI systems, autonomous platforms, and cybersecurity solutions for defense and national security.