Government Backlogs May Shrink Through AI, Not New Hires

A Different Workforce Equation

While private companies debate whether AI will replace workers, government agencies face a fundamentally different question: Can AI help them process work they currently cannot handle at all?

The private sector has begun using AI to avoid future hiring rather than eliminate existing jobs. IBM estimated in 2023 that roughly 30 percent of certain back-office functions could be automated over five years, and by 2025 reported that AI had taken over work previously done by a few hundred HR staff. Klarna initially claimed its AI assistant handled work equivalent to 700 full-time customer service agents, though the company reversed course in 2025 after its CEO acknowledged the AI-first approach had degraded service quality and began rehiring human agents.

But government is not starting from abundance. Public agencies already operate with unfilled positions, recruitment challenges, and budget constraints that make competing with private-sector salaries difficult. MissionSquare Research Institute's 2025 workforce survey found that recruiting challenges persist across state and local government, particularly in engineering, IT, skilled trades, and healthcare roles.

Why it matters

The bottleneck in government service delivery is capacity, not ambition. A planner retires and the city cannot replace her. A benefits office has open caseworker positions but no budget authority to fill them. A permitting department has applications waiting because review staff are overloaded. AI presents an option between hiring another person—often impossible—and letting work pile up indefinitely.

Throughput Over Replacement

The meaningful public-sector use case is not replacing public servants but increasing institutional throughput. Government work includes essential but low-judgment tasks: reading applications, checking missing documents, summarizing meetings, drafting routine letters, routing requests, and responding to common resident questions.

Agentic AI systems can read intake forms, identify missing documents, draft resident responses, check relevant policy, flag exceptions, create tasks for human reviewers, and update case records. The goal is removing avoidable drag around humans, not removing humans from decision-making.

Public agencies exercise judgment, interpret rules, and balance equity, law, risk, precedent, budget, and public trust. Those remain human responsibilities. But many agencies spend scarce human judgment on nonjudgment work—caseworkers hunting for documents across five systems, planners manually retyping information that already exists.

The Governance Gap

The dangerous version of this story involves agencies using AI as a budget workaround, cutting staff while pushing residents into automated systems that are difficult to understand or appeal. The labor does not disappear—it shifts from paid staff to residents, often those least able to absorb it.

AI can increase capacity and hide capacity failures simultaneously. It can reduce backlogs but also accelerate bad decisions. It can make services easier to access while creating new barriers for residents who need help, translation, or exceptions.

Agencies need clear rules for where AI can assist, where humans must decide, how residents can appeal, how errors are audited, and how workers are trained. This governance layer determines whether AI expands public capacity or weakens public accountability.

Vacancies as the Leading Indicator

The first workforce impact will likely appear through vacancies. A department with ten budgeted roles and seven filled positions may use AI to survive without filling all three. A permitting office may reduce review time without adding planners. A call center may use AI to answer routine questions because hiring bilingual agents is difficult.

This is still a workforce impact—it just looks like a job that never reopens rather than a layoff.

The agencies that succeed will start with workflow, not models. Where are residents waiting? Where are employees duplicating effort? Where does information get trapped? Where do rules require judgment? Then they will place AI around those constraints as infrastructure for public work, not as replacement for public servants.

These details were first reported by Government Technology.