Nine in ten public sector organisations operate across fragmented IT platforms. That baseline complexity is making AI adoption harder than the efficiency narrative suggests, according to new SolarWinds research covering 200 public sector IT professionals.
The numbers tell a consistent story: 56 percent of public sector IT workers say AI has made their roles more demanding; 74 percent say it is changing how work gets done without reducing how much there is to do. Both figures come from a SolarWinds survey of 1,040 IT professionals published today, with the 200 public sector respondents drawn from government, healthcare, and education.
The pattern emerging across the data is one of displacement rather than reduction. More than half (59%) say clearer AI policies and guardrails would help teams adapt. Cognitive load has increased for 32 percent, while a further 35 percent report it has reduced in some areas and risen in others. Skills gaps compound the pressure: the most in-demand competencies identified are designing AI-driven workflows (50%), evaluating AI outputs (43%), and interpreting AI-generated insights (42%).
Public sector organisations are under real pressure to do more with less, so it's understandable that Artificial Intelligence is being looked to as part of the answer. But, as powerful as it is, AI only helps if it's implemented thoughtfully, and if the technology itself doesn't become another burden for already stretched teams.
The fragmentation problem matters here. When 90 percent of public sector organisations say their IT systems span multiple disconnected platforms, AI tools layered on top of that estate inherit its complexity. Configuring and managing AI across fragmented environments takes people and time — precisely the resources the technology is supposed to free up.
Giblin put the practical test plainly: "If adopting AI requires a major project, complex configuration or dedicated resource to manage it, then the effort hasn't been reduced, it has just been moved. The tools that create the most value in resource-constrained environments are the ones that are practical to deploy, easy to use and able to deliver benefits from day one."
The SolarWinds data does not argue against AI adoption in the public sector. It suggests the current implementation approach — broad mandates for automation without adequate policy, training, or infrastructure consolidation — is producing outcomes that look a lot like the pre-AI workload, with an additional layer of governance on top.
