AI is widening the attack surface of the critical infrastructure facilities and services that state, local, and education (SLED) organizations run. In the final post of our Cybersecurity Awareness Month series, we explain how a whole-of-state security approach can help SLED agencies protect their critical infrastructure.

Key takeaways

1. Because cyber attackers prey on under-resourced targets, they often go after SLED organizations, especially those that operate critical infrastructure, including water systems, hospitals, mass transit, emergency services, and schools. 
2. In the latest research from the National Association of State Chief Information Officers (NASCIO), nearly 90% of state CIOs rated cyberattacks on critical systems as a high concern.
3. AI expands the attack surface in two fundamental ways: SLED agencies and employees adopt AI tools and AI-built apps faster than security teams can authorize and inventory them, while attackers use AI automation to find weak points faster.
4. A whole-of-state cybersecurity approach treats state agencies, counties, cities, school districts, and utilities as one defensive perimeter, with shared visibility, shared services, and clear governance.

Cyber attackers often look for the weakest link that can offer them an entry point into larger targets. If they can’t crack a state agency’s hardened systems, they’ll try to breach it in a roundabout way by first breaking into a small town’s water utility, a rural school district, or a county dispatch center, and then attempt to move laterally toward the bigger target. 

When one of those weak links fails, the consequences can reach well beyond the breached agency: halted 911 dispatches, disrupted water service, exposed resident data, and more.

AI is further complicating the picture. 

State, local, and education (SLED) organizations are adopting AI quickly to improve efficiency, employees are building applications with it informally, and adversaries are using it to speed up reconnaissance. Each of those developments is making it harder for security teams to protect SLED agencies.

In this fourth and final blog in our Cybersecurity Awareness Month series, we focus on what that expanding AI attack surface means for state, local, and education cybersecurity leaders responsible for critical infrastructure, and how whole-of-state cybersecurity can help them stay ahead.

Why SLED critical infrastructure is a prime target

SLED organizations operate critical infrastructure, such as water treatment plants, transportation systems, public hospitals, schools, and emergency communications, across many of the 16 critical infrastructure sectors categorized by the Cybersecurity and Infrastructure Security Agency (CISA). 

State technology leaders are concerned about protecting these critical infrastructure facilities and services from cyber attacks, as “The Expanding Cyber Perimeter” report, published in September by the National Association of State Chief Information Officers (NASCIO) and General Dynamics Information Technology (GDIT), found.

Here are five key findings from the report, which drew on the 2026 NASCIO State CIO Survey, the 2026 NASCIO-Deloitte Cybersecurity Study, and interviews with state CISOs:

• Nearly 90% of state CIOs rate cyber attacks on critical systems as a high concern. The rest rate them as moderate.
• State CIOs report low confidence in local governments’ cyber practices. The report’s authors also say it’s unclear how states should help local governments and independent local government entities, like special districts.
• Water and wastewater plants, dams, hospitals, and transportation services face the greatest cyber risk.
• Many systems run on aging or proprietary operational technology (OT) with no upgrade path, and broad remote access increases their cyber risk.
• Small communities and special districts often have limited cyber staff, aging equipment, and complex OT with few controls. With hundreds of such entities in a single state, attackers have plenty of entry points to try.

Responsibility is also murky. According to the NASCIO-GDIT report, it’s often unclear which agency is responsible for responding to an attack on critical infrastructure, especially where legal authority, governance, or partnerships aren’t in place.

Cyber risk extends beyond technical and geographical boundaries. When an attacker breaches systems of a county water utility or a school district, the incident can have widespread consequences. 

State and local officials and their CIOs and CISOs should stop treating these incidents as one city’s or one county’s problem. Robert Beach, then CTO of the City of Cocoa, Florida, and a member of the Multi-State Information Sharing and Analysis Center (MS-ISAC) executive committee, told StateScoop last year that an attack on critical infrastructure “is an attack on the nation.”

Insiders add another layer of risk. Another MS-ISAC report, “Strengthening Critical Infrastructure: SLTT Progress & Priorities,” published in August 2025, flagged excessive access to system controls as a vulnerability. The report recommended agencies review who can reach their networks and disable credentials promptly when employees leave. 

The 2026 NASCIO-Deloitte study underscores the risk, noting that attackers exploit compromised credentials and former employee accounts, so regular access reviews and prompt offboarding remain essential.

Recent cyber attacks against water systems have made the threat very real. On July 30, the Federal Bureau of Investigation (FBI) and the Environmental Protection Agency (EPA) warned that water and wastewater sector utilities in at least seven states had reported cyber attacks, leading in some cases to degraded water operations. 

That same day, the CISA said it had observed “a significant increase in cyber threat actors targeting programmable logic controllers” in the water and wastewater sector. 

“Threat actors targeting exposed PLCs have modified passwords to lock out operators and disconnected the PLCs by changing their IP addresses. This activity has resulted in boil water notices and sustained manual operations,” the CISA alert reads.

A week earlier, CISA and several federal agencies had published a joint advisory warning critical infrastructure organizations in the government, water, and energy sectors of “ongoing cyber exploitation of internet-connected OT devices.”

Tenable researchers recommend removing internet-exposed PLCs and setting their mode switches to “Run”, among other steps. The U.S. government is also tightening cybersecurity regulations for water and wastewater utilities. 

Federal and state leaders are mobilizing, too. 

In August, Texas Governor Greg Abbott, National Cyber Director Sean Cairncross, and Texas Cyber Command Chief Vice Admiral (Ret.) TJ White launched Project Watershed at a roundtable in San Antonio, Texas, which I was honored to attend. Project Watershed is a joint effort aimed at strengthening the cybersecurity of Texas water and wastewater systems against threats from hostile foreign adversaries. 

Governor Abbott said that through the project, “our communities and water systems will be better secured from the onslaught of cyber-attacks.”

Protecting the water systems Americans rely on every day is a shared responsibility, and Project Watershed shows the kind of public-private collaboration a whole-of-state approach depends on.

How AI expands the attack surface

AI has moved from pilot project to operational reality in state government. NASCIO’s 2026 research shows AI overtaking cybersecurity as the top policy priority for state CIOs, the first time in more than a decade that cybersecurity loses the top spot. 

AI adoption is good news for service delivery, but it also increases SLED agencies’ cyber risk. Three forces drive the rise in cyber risk:

1. Employees build and adopt AI tools on their own. As we covered earlier in this series, non-technical staff can now use AI tools to generate working applications using natural-language prompts, often without permission from the IT and security departments. In a government setting, those unsanctioned apps can access resident data, case management systems, and internal credentials, all without testing, patching, or monitoring.
2. AI shows up inside the software SLED agencies already own, as vendors quietly add AI features and connectors to their products, which can expand data access and permissions without the awareness of the IT and security teams.
3. Attackers use automation to move faster. The NASCIO-GDIT brief cites emerging threats such as automated reconnaissance, and it urges states to prepare for an increase in AI-enabled attacks on critical infrastructure, as the window between when an exposure appears and when attackers exploit it keeps shrinking.

For critical infrastructure operators, the stakes rise further. Their aging OT systems with broad remote access are already a soft target. Unvetted AI apps, over-permissioned connectors, and automated scanning by attackers quickly multiply the number of paths an attacker can chain together.

The key for SLED agency CISOs is to see their exposures across IT, OT, cloud, identity, and AI as one connected picture, and to prioritize and act on the ones an attacker could realistically exploit to reach systems that matter to public safety. An exposure management program can help SLED agencies address this cyber risk with comprehensive visibility, prioritization, and AI-assisted action.

What a whole-of-state approach looks like

A whole-of-state approach treats state agencies, local governments, schools, and utilities as one defensive perimeter instead of a collection of individual, siloed organizations. 

The state government should share governance frameworks, visibility of IT/OT assets and security findings, and cybersecurity capabilities so the smallest entity doesn’t have to defend itself alone. 

As Tenable’s Matt Kucik explains in the blog, “Whole-of-state cybersecurity: Uniting SLED agencies for maximum impact,” success starts with leadership and vision, open communication, and an actionable plan.

The model is gaining ground. The NASCIO-GDIT report found that 73% of states responding to the 2026 State CIO Survey include critical infrastructure cyber protection in their whole-of-state plans. It calls these models the main framework emerging for statewide resilience. Maturity varies widely, though. Four building blocks separate strong programs from weak ones:

1. Governance and clear roles. CISA’s state cybersecurity governance case studies highlight Georgia and Virginia, which use formal structures to treat cyber risk as a collective priority. They have proactively defined who is in charge of prevention, who is in charge of responding, and who is in charge of generating incident reports.
2. Shared visibility. You can’t prioritize what you can’t see. A statewide view of assets and exposures lets a small water district benefit from the same insight as a large state agency.
3. Shared services. In Utah, its federally funded shared services program offers endpoint protection, patching, training, and incident response to roughly 80% of local government entities in the state. The state secured an additional $1.5 million for water-sector cybersecurity.
4. Trust earned through results. What local partners need from the state is faster remediation, lower costs, and fewer surprises. New Jersey CISO Michael Geraghty made this point in the NASCIO-GDIT report: Telling local governments you’re from the state and here to help doesn’t earn their trust. Results do.

It’s also key to share information quickly. Terry Loftus, CIO of the San Diego County Office of Education and an MS-ISAC committee member, told StateScoop that many districts learn of attacks when they receive a ransomware demand. 

State cybersecurity agencies should build channels that proactively push threat intelligence and exposure data to every entity in their state, to increase the chances of preventing attacks.

Five steps SLED agencies can take to boost their cyber defenses

Here’s Tenable’s advice on where to start, whether you lead a state agency, a county IT shop, a school district, or a municipal utility:

1. Create and continuously update a complete inventory of your critical infrastructure assets, including OT systems, remote access paths, and AI tools and apps. Start with the systems that affect public health and safety.
2. Prioritize by exploitability and impact. Don’t chase every finding. Focus on the exposures an attacker could realistically chain together to reach water, power, health, or emergency systems. Removing internet-exposed controllers and securing any remote access paths to them should be the highest priority.
3. Share visibility across jurisdictions. Give small entities access to the same insight larger agencies have, whether through statewide tooling, shared services, or regular information-sharing forums.
4. Govern AI use before it sprawls. Set clear policies, approved-tool lists, and training for employees who build with AI.
5. Check if State and Local Cybersecurity Grant Program (SLCGP) funding is still available in your state to help fund these priorities.

How Tenable can help SLED agencies

Tenable One gives security teams unified visibility, insights and action across their organizations’ attack surface, including IT, cloud, identity, AI, and OT. 

Tenable One OT Exposure capabilities extend that visibility, insight, and action to the OT and extended internet of things (xIoT) devices used by many critical infrastructure organizations. 

With Tenable One, SLED teams can see what they have, understand which exposures matter most, and take prioritized actions, maximizing staff and grant dollars to drive down cyber risk. 

The City of Raleigh in North Carolina, for example, uses Tenable to ensure the safety and sustainability of its public water systems. To see how Tenable supports statewide programs, explore our state, local, and education solutions and learn how Tenable helps meet SLCGP requirements.

Looking ahead

The attack surface is growing, a trend that AI is accelerating. Federal funding is shrinking, and its future is uncertain. Yet the fundamentals of resilience stay the same: know your assets, understand your exposures, share what you learn, and act together before an attacker forces your hand.

No single state or jurisdiction has solved this, and no single tool will. A whole-of-state approach gives SLED organizations the best chance to stay ahead of the threats facing the critical services residents rely on every day.

Learn how Tenable helps SLED organizations implement statewide exposure management and align with SLCGP requirements.

Read the first three blogs in the Cybersecurity Awareness Month series:

• How to mitigate the risk from AI-generated apps built by your ‘citizen coder’ employees
• Critical infrastructure orgs face heightened cyber risk from nation-state groups, AI threats
• How to contain shadow AI, as it takes on multiple forms and spreads throughout your environment