Advertisement

AI tops cybersecurity in survey of state IT priorities

And “it wasn’t even close," said Doug Robinson, executive director of the National Association of State Chief Information Officers, the group that collected data on technology officials' top concerns.
Listen to this article
0:00
Learn more. This feature uses an automated voice, which may result in occasional errors in pronunciation, tone, or sentiment.
AI
(Getty Images)

In a development that will surprise few who’ve been tracking government, technology or world events more broadly, artificial intelligence has, on average, overtaken all other concerns in state technology offices across the nation.

The National Association of State Chief Information Officers, a group that represents state governments’ top IT officials, on Tuesday published its annual list of priorities, which is based on a survey of its membership. In last year’s list, AI was barely edged out for the top slot by cybersecurity, a technical and political concern that has made the list all 20 years since its first edition. But this year, said Doug Robinson, NASCIO’s executive director, AI won and “it wasn’t even close. It’s been a very swift ascension to No. 1.”

For state CIOs, generative AI tools represent a means of offsetting workforce shortages that have been slow to repopulate after the COVID-19 pandemic, chatbots to provide residents more personalized experiences when they visit government websites, or a scalpel to shave a few extra hours of work off processing applications for safety-net benefits that will receive less support under the second Donald Trump presidency.

In an annual membership survey published last October, NASCIO found that states were nearly ready to begin scaling up their AI projects. Ninety percent reported they were running AI pilot projects, 71% had begun training staff and 51% had created “sandboxes” for safely testing their creations. Only 25% reported having dedicated funding for AI, though outside organizations are offering at least some help there. The nonprofit Center for Civic Futures in November announced its AI Readiness Project, a $500,000 initiative designed to help states “move from curiosity to capability.” Last week, the center announced a separate $8.5 million round of funding designed to prove out a handful of government AI projects over the next two years.

Advertisement

Generative AI has absorbed persistent limelight since OpenAI unveiled ChatGPT to an astounded public in late 2022, but it took states a couple years of readying their task forces and crafting policy frameworks before they began nudging forward their first low-risk pilots or exploring uses beyond chatbots. Now, Robinson said, “it’s embedded in almost every tool now that states are bringing in.”

In 2023, NASCIO published a guide with 12 tips for states looking to do AI right. Robinson said the advice hasn’t changed: experiment, but also create strong governance — “Governance is probably top of the list along with data quality and remediating data and workforce readiness.”

If AI topping NASCIO’s list was an obvious development, “budget and cost control” jumping up to the No. 3 spot might be an ominous one. Robinson called it “a realistic sign” of where states are headed. Recent research by the National Association of State Budget Officers shows that state revenues are flattening out after enjoying unexpected growth during the pandemic. Some state technology officials have said they’re not worried about their budgets, but many CIOs have expressed varying degrees of concern.

“It’s the reality of what the states are seeing and the CIOs realize that,” Robinson said. “The warning flags are up in terms of revenue shortfalls. … CIOs are hearing from agency customers that their budgets are going to be tapped.”

In states facing budget deficits or revenue declines — and 23 states are projecting general fund spending to decline, according to NASBO — CIOs can feel the pain twice: once when the agency customers they bill have their funding reduced and again when their own funding is reduced.

Advertisement

“They get hit from both ends,” Robinson said.

AI could prove useful to states hoping to maintain current levels of service without receiving funding to hire additional staff, but such technological developments take time. More widespread adoption of AI-powered productivity tools across state workforces are proving modestly efficient, and Pennsylvania’s governor has even claimed that bringing ChatGPT to staff was saving them an average of eight hours of work each week. But larger projects — designed to test out innovative uses of AI, scale them across state governments and then export them to other states — could take years.

And for some projects, like improving digital accessibility, states don’t have years. States (and large cities) have until April until a Department of Justice deadline requires accessibility compliance on their websites and other digital properties. Some states are making rapid progress, but many are not expected to make the deadline. On NASCIO’s list, accessibility jumped to the No. 6 slot, up from No. 10 last year.

“[CIOs] are feeling the sense of urgency there,” Robinson said. “It is not their responsibility directly. It’s the responsibility of the agencies, but they have overarching enterprise responsibility to make that move forward.”

NASCIO’s most recent annual survey found that two-thirds of states have hired a technology accessibility coordinator, but only 46% have funding for accessibility improvements. Only one state had fully implemented its plan. Some states are turning to specialized software, sometimes powered by AI and other automation technologies, to better inventory their web assets, though the work of updating websites and the tens of thousands of PDFs they host remains, by many accounts, at least partially a tediously manual process.

Advertisement

“They’ve been kicking the can down the road a long time and now they’ve got a federal compliance mandate,” Robinson said.

The bottom of the list contains items that are still important for many states — cloud infrastructure (No. 10), consolidation and optimization (No. 9) and enterprise architecture (not on the list, but at No. 11, according to Robinson). But a decade ago or less, these concerns rivaled cybersecurity in importance. Robinson said their decline is a mark of maturity in state governments — “It’s become part of the standard operating model.”

Latest Podcasts