Government audit, characterized by systematic examinations and evaluations of public sector organizations, is pivotal in ensuring the integrity, transparency, and accountability of public finances. Given this scenario, this paper presents a comprehensive exploration of the integration of Machine Learning (ML) within Government Auditing, with a specific focus on financial statements, and to study how ML has influenced the analysis and decision-making of the auditors. Following the PRISMA methodology, this literature review explored publications from the past five years using keywords in both English and Portuguese to ensure comprehensive coverage. The review revealed the multifaceted applications of ML in various contexts of financial auditing within governments worldwide, elucidating the diverse approaches employed in these scenarios and their success rates, providing valuable insights into emerging trends and best practices. Overall, this paper contributes to advancing knowledge and understanding of ML in Government Auditing, providing valuable insights for practitioners, policymakers, and researchers in the field.