This paper investigates the impact of Artificial Intelligence (AI) and Machine Learning (ML) for enhanced fraud detection in the U.S. banking industry. As electronic transactions increase, rule-based approaches are less effective in identifying sophisticated cyber threats. This research investigates how artificial intelligence and machine learning technologies are enhancing fraud prevention, real-time monitoring, behavioral analysis, and anomaly detection in banking systems. The study also examines these implementations within strict federal regulatory frameworks, including the Gramm-Leach-Bliley Act (GLBA), Sarbanes-Oxley Act (SOX), and Federal Financial Institutions Examination Council (FFIEC) guidelines. The research addresses significant implementation challenges that financial institutions encounter when deploying AI/ML systems. These challenges include data privacy concerns, algorithmic bias mitigation, and the critical requirement for model explainability to satisfy regulatory requirements and maintain stakeholder confidence. The paper examines machine learning models such as supervised, unsupervised, and ensemble learning, and their role in fraud detection, insider threat, and systemic risks. Thus, the analysis provides a comprehensive operational and regulatory template to help banks in the United States respond to the adoption of AI/ML-based fraud detection systems and to strike a balance between technological progress and legal and financial system soundness.