The federal government’s ongoing assessment of artificial intelligence (AI) applications highlights a significant surge in adoption, with agencies across the United States cataloging over 2,500 use cases. This marks a considerable increase from previous inventories, signaling a shift from limited pilot programs to operational systems now embedded in daily government functions. The latest federal inventory, publicly reported in 2024, detailed 2,133 use cases across 41 federal agencies, according to FedScoop’s coverage of the Office of Management and Budget’s annual tally. As more use cases transition from pilot to production and additional agencies broaden their disclosures, updated totals are expected to rise again this year.
Federal Chief Information Officer Greg Barbaccia noted that the annual inventory not only aims to track AI adoption but also to standardize governance and oversight as these technologies expand across various agencies. In a survey conducted by Google Public Sector, nearly 90% of federal IT leaders indicated that their agencies are either planning to implement or are already utilizing AI. This data suggests that the scope of AI adoption within government settings is broader than commonly perceived.
The most prevalent use cases identified include document and data processing, workflow and process automation, and decision support systems. Agencies are increasingly deploying AI to analyze extensive datasets, monitor network activities in real-time, streamline case management, and bolster fraud detection efforts. Many of these applications are focused on areas where data volumes are substantial and manual review processes require significant resources.
At the State Department, the integration of AI has necessitated a fundamental shift in how technology is introduced to the workforce. Chief Information Officer Kelly Fletcher described a departure from the traditional IT rollout model. “We are actually doing something a little bit different with AI than what we’ve done with IT historically,” Fletcher stated. “Historically, we would roll out technology, we would test it really carefully, and then it would be sort of like etched in stone.”
In contrast, the State Department launched its generative AI chatbot, known as State Chat, to an initial group of 300 users over a year and a half ago. This early version was designed to be iterative, with user feedback shaping its development. The pilot group eventually expanded to around 3,000 users who provided continuous feedback, allowing for prompt changes and feature updates that users often noticed before analytics reflected these shifts. “These users both affect what are the next capabilities that we add to the chatbot, but then also identify problems,” Fletcher explained, characterizing a feedback-driven model that treats AI as an evolving capability rather than a static product.
Meanwhile, the U.S. Army faces distinct challenges related to workforce readiness as it adapts to the rapidly changing landscape of AI. Kris Saling, Chief Technology Officer for the Office of the Assistant Secretary of the Army for Manpower and Reserve Affairs, remarked that the Army’s previous AI strategies were established before the rapid emergence of generative AI tools. “Strategy was written back in 2018 for the most part for the Army,” Saling noted, referring to the game-changing nature of advancements like ChatGPT and the swift proliferation of generative AI technologies.
The Army has successfully figured out how to proliferate AI rapidly, Saling acknowledged, but the challenge remains in training personnel to utilize these tools responsibly. He emphasized that the issue extends beyond technical training to include ensuring that personnel comprehend the domain context in which AI systems operate. The Army is working to unify previously siloed systems and strategies into a more cohesive approach.
Despite the growing number of federal AI use cases, challenges persist. The Google Public Sector survey identified security and adversarial risks as the primary barrier to scaling AI, cited by 48% of respondents. Concerns over reliability followed closely at 35%, while workforce-related issues, including potential disruptions and skills gaps, were also highlighted by agencies navigating large-scale deployments.
As AI continues to reshape the landscape of government operations, the increasing adoption underscores a broader transformation within federal agencies. This evolution not only reflects the potential for enhanced efficiency and effectiveness in service delivery but also raises critical questions about governance, ethical implications, and the future readiness of the workforce in a rapidly advancing technological environment.
See also
OpenAI’s Joseph Larson: Infrastructure Essential for Effective Public Sector AI Use
Trump and Hegseth Target Anthropic’s AI Guardrails, Advocating for Unrestricted Development
Senate Passes AI Bill of Rights, Mandates Parental Consent for Minor Usage
UK’s £27M AI Skills Program Falls Short; 56% of CEOs See No ROI Yet
Israel’s Cyber Chief Yossi Karadi Reveals Key Strategies to Combat AI-Driven Cyber Threats
