AI Oversight in US Government: From Formal Policies to Functional Accountability

Abstract

When artificial intelligence (AI) gained attention about ten years ago, its application in federal agencies has increased dramatically. Today, it is a vital tool for efficiency, security, and creativity across federal agencies, having begun as an experimental technology with few government applications. However, a big disconnect exists between stated policies and real supervision procedures due to the US government agencies' expanding usage of AI systems. By examining federal regulations, auditing instruments, and agency-specific procedures, this study seeks to determine how successful the US government's present AI supervision procedures are. The findings of this paper show that, although some agencies have been able to put the overview of artificial intelligence in place, other small agencies that do not have enough resources to fund the project or lack trained personnel to carry out the oversight lag behind. Drawing from case studies, policy documents, and academic research, this study spotlights challenges in guaranteeing accountability, weaknesses in enforcement, and limitations of the existing auditing processes. The research aims to support the creation and development of functional oversight mechanisms that fully preserve and safeguard the interests of the public while maintaining democratic ideals. This has been accomplished by developing tactics and best practices to increase accountability. By informing stakeholders and policymakers about the necessary adjustments to bridge the gap between AI governance practice and policy, the findings of this research will ultimately boost public trust in government AI applications.