Agencies are about to make the same mistake with artificial intelligence that they made with cloud, and this time it will cost them much more than a missed migration deadline. There’s a massive mismatch between the processes of the federal bureaucracy and the speed at which AI operates. For instance, if a program office adopts AI-assisted analysis which cuts research time from weeks to days, those gains disappear when it must route every finding through a review board that still meets monthly.
It’s this workflow and process mismatch that’s the primary bottleneck, but the federal government is approaching AI just as it does every large-scale technology shift by reducing it to a training line item. Cloud migration became “cloud certifications.” Cybersecurity modernization became “security awareness modules.” Now AI is following the same script. Agencies are standing up AI training programs, checking the workforce readiness box and waiting for adoption to follow.
It won’t, but not because the quality of instruction is poor, but because training people on discrete IT tools is a shortsighted strategy. AI is not a tool that easily slots into the way agencies already work because it changes the nature of the work itself, altering how decisions get made, how fast information moves, and how humans and systems divide labor. What’s more, given how quickly AI tools are advancing, they’re changing faster than any curriculum can keep up. An employee trained on a specific AI system this quarter may be working with a materially different tool six months later. Training for tool-specific proficiency inherently has a short shelf life.
Who owns the decision when AI collapses the timeline?
Instead, the federal government needs to create teams and larger organizations that can adapt quickly, reorganizing work around new capabilities as they emerge. Most federal roles were designed around a linear model of work. Requirements are defined, tasks are assigned to individuals or teams, and value is delivered at the end of a fixed cycle. AI breaks that linearity, because it can compress weeks of human-directed research, drafting and analysis into a task that takes hours. That said, an organization will only benefit from this capability if the surrounding roles, approval chains and funding cycles can move at the same speed.
Right now, most agencies cannot do this. A program office might adopt AI-assisted analysis, but still route findings through a review structure built for a much slower cadence, so the bottleneck moves from the process of researching and creating to oversight. As a result, workforce modernization in the federal government must include role redesign to clarify who owns decisions when AI shortens the distance between insight and action. The judgment and authority to act on AI insights must be built into roles that didn’t previously require it.
Leaders need to decide faster, not just know more
AI training programs tend to focus on end users, but the leaders sitting above those users are the ones who need the most significant shift in how they operate. With AI tools generating options, drafts and analyses in near real time, the constraint on mission delivery moves from “How fast can we produce information?” to “How fast can leadership evaluate it and decide what matters most?” This shift requires leaders who are comfortable making higher-velocity, iterative decisions rather than waiting for a fully resolved picture.
What’s more, government planning and funding cycles were not built for the rapid pace of innovation that AI enables. Multi-year budget requests and annual planning cadences all assume a relatively stable understanding of the work ahead. AI shortens delivery cycles and surfaces new possibilities mid-stream, which makes that assumption of stability harder to sustain.
Agencies that have made real progress on agile and lean-agile portfolio management already have a head start in moving to a new AI-native model. These agencies fund value streams instead of individual projects, and they plan in shorter, adaptive increments. As a result, these organizations now have the flexibility to redirect effort as AI changes what’s achievable, without waiting for the next budget cycle to catch up. Agencies still operating on rigid, waterfall-style planning will find that even excellent AI training can’t compensate for a planning structure that can’t absorb the pace of change AI creates.
What actually moves the needle
None of this means training has no place. People do need a working understanding of what these tools can and can’t do. But training is the floor, not the strategy. Turning AI investment into mission results requires organizations to evolve how they lead, govern, fund, plan and deliver work. Specifically, agencies need to look at four aspects of their organizational structure:
- The roles people hold.
- The pace at which leaders are equipped to decide.
- The way humans and AI systems divide day-to-day work.
- The planning and funding structures that determine how quickly the prior three areas can change.
Agencies that treat this as a workforce-training initiative will end up with employees who know how to use AI tools inside an organization that still can’t move at AI speed. Agencies that treat it as an operating model problem that touches roles, leadership behavior, and how work gets planned and resourced will be positioned to turn AI investment into mission outcomes. AI won’t wait for the next budget cycle. Neither can the workforce built to use it.
Travis Moorer is strategic advisor for organizational agility, digital transformation and AI at Scaled Agile and a U.S. Army Veteran.
Copyright © 2026 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.