The debate around AI in government often focuses on whether agencies have the political will to change or the budgets to invest. Justin Fulcher shifts the focus to a more specific set of obstacles: siloed data, outdated processes, and compliance frameworks built for workflows that no longer exist.
These structural problems, Fulcher argues, are what actually determine whether AI investments deliver results or simply add to the complexity of already burdened agencies. Understanding them clearly is the starting point for any productive conversation about modernization.
The Anatomy of Institutional Drag
Justin Fulcher uses the phrase “institutional drag” to describe something concrete. Government agencies routinely operate with data spread across systems that can’t communicate with each other. Decisions that could be automated still require manual review steps because the compliance framework says so, even when those requirements were designed for paper-based operations. Procurement processes that could move in weeks take years because the rules governing them were written before modern software existed.
“Across government, healthcare, defense, and infrastructure, our core systems operate as if it were 1975,” Justin Fulcher wrote in a piece on institutional renewal. That framing captures both the scale of the problem and its nature it isn’t a technology gap, it’s a process gap.
Where AI Can Actually Help
The implication of Fulcher’s analysis is that AI investments should target friction reduction at the process level. Tools that automate repetitive decision-making steps, connect data across previously siloed systems, or reduce the manual overhead in compliance review can address the actual drag points.
Tools that are grafted onto existing broken processes without addressing the underlying workflow problems tend to fail or, at best, produce limited gains. During his time at the Department of Defense, Justin Fulcher contributed to acquisition reforms that cut software procurement timelines from years to months. That work was built on understanding the specific process bottlenecks first. The lesson for agencies deploying AI today is direct: understand the friction before deploying the fix. Justin Fulcher’s experience suggests the diagnosis matters as much as the technology. Visit this page on LinkedIn, for more information.
Find more about Justin Fulcher on https://www.facebook.com/JustinLFulcher/