Justin Fulcher on Building AI Systems Government Workers Will Actually Use
Technology deployments in government fail for many reasons. Procurement delays, budget overruns, and technical mismatches all contribute. But one of the most common, and least discussed, failure modes is simpler: the people who are supposed to use a new system do not trust it or find it makes their work harder rather than easier. Technology founder and former Defense Department advisor Justin Fulcher has argued that building AI systems government workers will actually use requires understanding that problem and designing for it from the start.
Adoption Is the Real Benchmark
Fulcher’s framework for evaluating AI in government does not focus primarily on what a system can theoretically accomplish. It focuses on whether the system will earn adoption. That distinction matters because government agencies have a long history of technology deployments that worked in pilots and failed in practice, often because they were designed around what administrators wanted rather than what the workforce could absorb.
Justin Fulcher has pointed to a clear principle: technology adoption in regulated environments succeeds when it reduces existing friction rather than creating new complexity. AI tools that require extensive retraining, generate new compliance concerns, or depend on integrations that agencies must build from scratch will face resistance regardless of their technical merits. Tools that fit cleanly into existing workflows and deliver immediate, visible time savings will gain traction.
The applications that meet that test in most federal environments include document processing, data synthesis, routine compliance verification, scheduling coordination, and correspondence management. These are areas where the volume of work is high, the tasks are well-defined, and the efficiency gains are easy to demonstrate. Justin Fulcher has argued that this is where agencies should focus their initial AI investments, building trust through demonstrated results before attempting more complex deployments.
The Workforce Trust Problem
AI deployment in government faces a workforce trust challenge that private-sector implementation often do not. Federal employees operate under civil service protections and work within institutional cultures that have absorbed many rounds of technology-driven reform promises. Skepticism is not irrational; it is the product of experience.
Fulcher draws on his background at RingMD, where the company-built healthcare technology across highly regulated Asian markets, and at the Department of Defense, where he contributed to acquisition reform and IT modernization efforts that reduced procurement timelines from years to months. In both environments, earning workforce trust was prerequisite to achieving durable results.
That experience shapes how he thinks about AI accountability. “Critical work is defined less by certainty at the outset than by stewardship over time,” Fulcher has noted. Systems must be auditable and explainable. They must fail safely when they encounter edge cases. And they must be maintained and improved based on the feedback of the people using them. Justin Fulcher’s consistent message is that AI’s long-term contribution to government modernization depends not on the sophistication of the technology but on the discipline of the implementation. Refer to this article for related information.
More about Justin Fulcher https://x.com/JustinFulcher
