The integration of artificial intelligence (AI) into the vast and complex machinery of the United States federal government is accelerating, yet it remains an uneven and cautious process. The appetite for AI adoption across federal agencies spans a spectrum driven by the technology’s promise to transform public service efficiency and effectiveness. At its core, AI offers the potential to automate routine workflows, extract nuanced insights from complex data, and revolutionize mission-critical functions ranging from health care to law enforcement. Despite these possibilities, significant institutional and operational barriers continue to slow widespread adoption.
Over the past decade, American administrations—regardless of political affiliation—have recognized the strategic imperative of embedding AI into government processes. Initiatives under both the Trump and Biden administrations have underscored AI’s role in enhancing service delivery, advancing data-driven policymaking, and maintaining national security and competitive advantage. Yet, this policy-level enthusiasm contrasts with the realities on the ground, where AI implementation is concentrated within a handful of large agencies, while many smaller entities remain on the sidelines.
Recent analysis from the Brookings Institution, based on federal datasets and specialists’ interviews across eight federal agencies, reveals compelling trends and challenges. From 2023 to 2025, reported AI activity nearly quintupled, with agencies expanding projects that employ machine learning algorithms, natural language processing, and automated decision-making systems. By 2025, AI-powered projects numbered over 3,600 across 41 agencies, illustrating increased reliance on AI for improving back-office efficiencies as well as frontline applications. However, the distribution remains highly skewed, as over half of all AI activity is concentrated among just five agencies, all large-scale organizations with extensive resources.
This disparity highlights a critical structural issue: smaller and mid-size agencies face disproportionate hurdles in accessing AI technologies. Constraints such as outdated infrastructure—legacy computing systems and fragmented data ecosystems—thwart AI’s potential to scale. Moreover, agencies working with sensitive health or security data navigate stringent compliance requirements and ethical concerns, limiting the scope for experimentation. These factors perpetuate a digital divide within the federal government itself, hampering the uniform rollout of AI capabilities.
Another pressing bottleneck is the federal workforce challenge. The bureaucratic inertia of government hiring practices contrasts starkly with the dynamic demand for AI talent. Lengthy recruitment timelines and a lack of clear career progression paths for technologists discourage top AI professionals from entering or remaining within federal employment. Initiatives by the executive branch seek to expedite hiring and enhance workforce development, but systemic reforms lag behind. Compounding these difficulties, recent workforce reductions—including layoffs of approximately 300,000 federal employees—have precipitated a precipitous drop in AI-focused job openings, further impeding AI adoption momentum.
Institutional culture also plays a formidable role in shaping AI’s trajectory within government operations. The ingrained risk-averse mindset prevalent in federal agencies manifests as resistance to adopting innovative and experimental approaches, particularly when AI’s internal decision-making processes remain opaque. The “black box” nature of many AI models generates skepticism among policymakers and implementers, especially concerning fairness, accountability, and transparency. Such mistrust limits AI deployment in domains where reliability and ethical considerations are paramount.
A politicized environment surrounding AI technologies further complicates adoption. Large language models (LLMs), for example, have come under scrutiny for their perceived biases and potential security risks. The label of “supply chain risk” applied to Anthropic’s Claude by the Department of Defense exemplifies growing concerns about dependencies on third-party AI providers and the geopolitical implications of AI sourcing. Similarly, AI systems like Grok, developed by Elon Musk’s xAI, have faced criticism for embedding controversial perspectives, illustrating how political dimensions can influence perceptions and acceptance of AI tools.
Despite these hurdles, Brookings fellow Valerie Wirtschafter finds cause for optimism, highlighting AI’s transformative potential when judiciously applied. Her research suggests that targeted interventions could accelerate AI’s integration into federal missions, creating tangible benefits for public services. These include streamlining bureaucratic processes, optimizing resource allocation, and improving service outcomes in health care, social benefits, and national safety.
To move forward effectively, the federal government must enhance its talent acquisition and retention strategies, making AI expertise a recognized and rewarded career path. This requires not only streamlining hiring protocols but also investing in ongoing professional development and AI literacy across the workforce. Cultivating an environment where staff understand AI’s capabilities and limitations is crucial for informed adoption and governance.
Transparency and knowledge-sharing across agencies can also play a vital role. Documenting successful AI implementations and disseminating best practices reduces redundancy and helps overcome skepticism by demonstrating empirical value. By fostering open communication channels, agencies can collaboratively tackle common challenges and refine AI deployments in alignment with ethical standards and public expectations.
Lastly, investments in AI should prioritize high-impact applications that directly improve citizen experiences and outcomes. Rather than expending resources on diffuse or low-priority projects, federal AI strategies should emphasize initiatives with measurable benefits, such as accelerating benefits delivery, enhancing medical diagnostics, or augmenting law enforcement capabilities with bias-mitigated AI tools.
In sum, while AI adoption within the federal government has advanced considerably in recent years, it remains a nuanced and challenging endeavor shaped by organizational culture, workforce dynamics, technological complexity, and political context. Continued research, policy innovation, and pragmatic implementation strategies will be essential to harness AI’s full promise in public sector governance, helping the U.S. government transform its operations in an increasingly digital world.
Subject of Research: People
Article Title: Assessing the state of AI adoption across the federal government
Web References: https://www.brookings.edu/articles/assessing-the-state-of-ai-adoption-across-the-federal-government/
Keywords: Artificial intelligence, Federal government, AI adoption, Workforce challenges, AI governance, Machine learning, Public sector innovation, AI infrastructure, Risk-averse culture, Large language models, AI transparency, National security
