Justin Fulcher Identifies the Highest-Value AI Applications for Federal Workflows
By Space Coast Daily // April 1, 2026
The question of where artificial intelligence can do the most good inside the federal government has become one of the more pressing debates in technology policy. For Justin Fulcher, a technology entrepreneur and former Senior Advisor to the Secretary of Defense, the answer begins with a clear-eyed distinction: deploy AI where it handles information, not where it controls operations.
“The highest-value applications are in areas that involve large volumes of information rather than direct operational control,” Fulcher said in a recent interview. “Federal agencies spend enormous time reviewing documents, analyzing data, and navigating administrative processes. AI can dramatically accelerate tasks like document analysis, regulatory review, procurement workflows, and internal research without introducing significant operational risk.”
Where Government’s Workload Creates an Opening
The scope of the problem Fulcher describes is not abstract. Federal agencies process millions of documents annually across regulatory, legal, and procurement functions, many of which still move through manual review cycles that predate modern computing norms. According to Gallup, 43% of public-sector employees reported using AI at least a few times in Q4 2025, up from just 17% in Q2 2023, a trajectory that reflects growing recognition of the technology’s practical utility. Yet adoption in government still trails the pace seen in leading private-sector industries, and the gap is widest in precisely the information-intensive workflows Fulcher points to as the most promising targets.
His reasoning tracks with broader productivity research. A Federal Reserve Bank of St. Louis survey found that generative AI users reported average time savings of 5.4% of their work hours, equivalent to roughly 2.2 hours per week for a full-time employee. For agencies handling thousands of staff hours devoted to document-heavy administrative work, even modest efficiency gains at scale represent a substantial operational shift.
Fulcher frames these applications as a natural fit because they sidestep the accountability concerns that make AI adoption in government particularly fraught. Document analysis, internal research, and procurement workflow support do not require machines to make autonomous decisions. They require machines to process and surface information faster than humans can do alone.
Augmentation, Not Automation
The second category Fulcher identifies carries equal weight. Decision support, he argues, is where AI can meaningfully improve the quality of analysis without removing human judgment from the equation.
“Another major opportunity is in decision support, which involves helping analysts synthesize large datasets, identify patterns, and surface relevant information more quickly,” Fulcher said. “In those cases, AI functions as an augmentation tool rather than an autonomous decision-maker.”
That distinction matters in a government context. Federal agencies operate under legal and oversight frameworks that require human accountability for consequential decisions. An AI system that helps an analyst identify a pattern in procurement data is fundamentally different from one that approves or denies a contract. Fulcher’s framing keeps the two categories separate, which is precisely why he views augmentation tools as lower-risk entry points for agencies still working through how to govern AI adoption.
The Accountability Principle
Across both categories, Fulcher returns to the same core principle. “The key principle is keeping humans firmly in the loop while using AI to reduce friction in information-heavy workflows,” he said. “Done correctly, that approach can significantly improve government productivity while maintaining accountability and oversight.”
Justin Fulcher’s perspective carries weight in part because of where it was formed. During his tenure as a Senior Advisor to the Secretary of War, he worked on acquisition reform initiatives that contributed to shortening software procurement timelines from years to months. That experience gave him direct exposure to the institutional friction that slows technology adoption inside the federal bureaucracy.
His view of AI in government is not that the technology requires a cultural revolution to take hold. It requires a practical one. Agencies that start by automating the right category of work, information processing over decision-making, can build institutional confidence in AI without assuming the risks that have historically made government technology modernization so contentious.
