GUEST COLUMN.

Artificial intelligence no longer feels like a niche IT topic. As we enter 2026, it continues to show up in nearly every product or service we touch—new laptops and software suites, cars, appliances, and “smart” home devices. For state and local government technology leaders, AI remains the headline issue: how do we encourage practical experimentation while establishing sensible guardrails, policies, and training?

Two years ago, at a local government CIO gathering, several participants used a phrase that captured the mood: “AI fatigue.” That reaction may sound surprising—today’s generative and agentive AI boom is only a few years old—but the feeling lingers to this day. AI weariness is the exhaustion and sense of being overwhelmed that individuals and organizations experience as AI tools and offerings proliferate, expectations rise, and pressure to “do more with AI” collides with constrained budgets and limited staff capacity.

It is amplified by the gap between ambitious promises and the actual results delivered in government applications. Years later, weariness lingers.

Government IT organizations live with a tension that many private firms can sometimes avoid: leaders are expected to adopt modern capabilities quickly while sustaining mission-critical operations and delivering dependable citizen services. Scaling AI introduces hurdles that drain energy and confidence—high costs and risks, heavy demands for data preparation and management, the need to redesign workflows and roles, rapid product churn, and the ongoing burden of training and experimentation.

How AI weariness shows up

AI weariness is not one thing; it’s a bundle of pressures that appears in different ways across an organization:

·        Cognitive overload: mental exhaustion from constant interaction with AI-enabled tools and repeated workflow changes.

·        Technological overload: being swamped by platforms, vendors, pilots, and updates—leading to decision paralysis or resistance.

·        Ethical and emotional strain: discomfort or distrust when AI affects human outcomes, raising concerns about fairness, privacy, and accountability.

·        Burnout and anxiety: fatigue among those asked to build, procure, integrate, and monitor AI—paired with fears of job displacement, rising workload expectations, and constant change.

The result is a common question among many leaders: why is this still happening—and what can we do about it?

Other technology shifts—cloud computing, broadband proliferation, and the rise of digital government—created stress, but they did not commonly produce a “fatigue” narrative. AI is different for two reasons.

First, adoption has been extraordinarily fast. Generative AI moved from novelty to mainstream almost overnight, creating a constant sense of falling behind. And today, we are learning more about Agentive AI.

Second, AI feels more personal than earlier technologies. It emulates human productivity and creativity and can supplement repetitive work at an unimaginably high speed. That speed is impressive—and unnerving. When outputs arrive instantly, people may treat them as “answers,” even when they can be incomplete, biased, or simply wrong. Public-sector leaders worry that flawed output could cause unintended harm to residents. They also worry about data integrity, bias, and whether AI is slipping outside of IT management’s control as underlying algorithms evolve.

In short, AI’s pace leaves public leaders with less time to think, plan, and build the organizational muscle needed to adopt it effectively.

The remedy: turning weariness into managed momentum

The first step is naming the experience. Many IT leaders and staff are feeling the same pressure and acknowledge that reality reduces isolation and defensiveness.

From there, governments can reduce weariness by shifting from “rush to deploy” to “disciplined adoption”:

1.     Start small and scale deliberately. Select a few high-value, lower-risk use cases (often internal first), set success measures, and define oversight checkpoints and stop rules before expanding.

2.     Train for the work people actually do. Beyond awareness sessions, staff need practical guidance on writing prompts, validating outputs, protecting sensitive data, and knowing when not to use AI.

3.     Involve employees early. Co-design workflows with the people who do the work. Emphasize augmentation over replacement and discuss openly how roles may change.

4.     Build clear guardrails and peer networks. Ethical guidelines and usage policies must be written in plain language, consistent, and reinforced by leadership behavior. Just as important, create communities of practice—inside your agency and across governments—where staff can share lessons and templates.

A final perspective: AI has been researched for decades, but modern generative AI remains (and now agentive AI) in its infancy. It can already “converse,” and unlike a human infant, it never sleeps—creating the perception that leaders must respond around the clock. Recognizing AI weariness, watching for its symptoms, and leaning into disciplined governance and human networks may be the best way to sustain momentum without burning out the people tasked with making AI work for the public good.

The contents of this Guest Column are those of the author, and not necessarily Barrett and Greene, Inc. 

#StateandLocalTechnologyManagement #ArtificialIntelligenceWearinessAndStateLocalGovernment #StateandLocalArtificialIntellligenceFatigue #ArtificialIntelligenceWeariness #ArtificialIntelligenceAndEmployeeBurnout #StateandLocalAI #StateandLocalArtificialIntelligence #StateandLocalGovernmentGenerativeAIPolicyAndManagement #StateandLocalGovernmentManagement #StateandLocalPerformanceManagement #StateandLocalTechnologyPerformance #CityGovernmentTechnologyManagement #CityPerformanceManagement #CityTechnologyPerformance #CityArtificialIntelligencePerformance #CityArtificialIntelligenceWeariness #CountyArtificialIntelligenceWeariness #StateLocalGovernmentGenerativeAIPolicyAndManagement #NationalAcademyOfPublicAdministration #NAPAStandingPanelOnTechnologyLeadership #StateandLocalAITraining #StateandLocalAIGuradrails #EmployeeEngagementAndArtificialIntelligence #ArtificialIntelligenceResults #ArtificialIntelligenceEvaluation #ArtificialIntelligenceAndWorkplaceFatigue #ArtificialIntelligenceAndManagedMomentum #AlanShark #GeorgeMasonUniversityScharSchool #StateandLocalEmployeesAndTechnologyOverload #BandGGuestColumn #BarrettandGreeneInc