While nearly 80% of public sector information technology leaders say their government’s AI adoption is aligning with AI oversight planning, just 37% have a formal, actively enforced AI governance framework in place according to a recent survey by research firm UserEvidence.
Only 19% of state and local government IT professionals reported an active AI oversight framework, and just 20% say such a framework exists at all.
"Confidence in AI governance is running well ahead of the actual controls agencies have in place, and that gap is where risk quietly accumulates,” Travis Galloway, senior advisor, government affairs at SolarWinds, a software provider that commissioned the report, said in a news release.
The survey of more than 800 public sector IT professionals in the U.S. and U.K. found less than 40% of respondents reported widespread AI use within their organization. However, AI interest and exploration is high, with only 5% of respondents saying that no AI tools are in use or in planning stages.
“The urgency around AI adoption is pushing public sector IT teams way outside of that comfort zone, before the right foundations are in place,” the report states.
Cybersecurity incidents in the public sector are on the rise. However, 35% of respondents said they are loosening security or compliance restrictions to enable AI access, the survey found. Nearly 45% of respondents reported an increase in AI-related security incidents over the past year.
Twenty-four percent of U.S. respondents have fully implemented zero trust architecture, a cybersecurity framework designed to enhance verification measures, and 36% report their framework is “mostly” implemented.
AI adoption varies between public sector segments. Thirty-two percent of federal government IT leaders reported “aggressively pursuing” AI adoption, while 12% of state and local government IT professionals reported the same. Sixty-eight percent of state and local respondents said they were pursuing AI adoption “methodically,” 17% said their approach was “hesitant” and 4% were “skeptical.”
Generative AI tools dominate the AI types survey respondents use, with 77% reporting their employer is using or piloting genAI tools such as Microsoft Copilot, ChatGPT and Google Gemini. More than 50% reported using AI-assisted cybersecurity tools, 49% deploy AI-assisted IT monitoring or alerting and 46% use AI-assisted help desk or IT service management.
Thirty-five percent of respondents said their employers are using or piloting agentic AI tools, which can take action without human input.
Fewer than 40% of respondents reported that AI tools are fully integrated into their standard monitoring operations.
