Critical utility systems face unprecedented cybersecurity challenges as digital transformation accelerates across energy, water, and telecommunications infrastructure. This study examines the integration of artificial intelligence (AI) and machine learning (ML) technologies into cyber risk management frameworks for critical utility systems. Through a comprehensive analysis of current literature, industry case studies, and experimental validation, we demonstrate that AI/ML-enhanced cyber risk management systems can reduce threat detection time by 73% and improve incident response effectiveness by 68% compared to traditional approaches. Our findings reveal that predictive analytics, anomaly detection, and automated response mechanisms significantly enhance the resilience of critical infrastructure against sophisticated cyber threats. The research provides a roadmap for utility operators to implement AI-driven cybersecurity strategies while addressing key challenges including data quality, algorithm bias, and regulatory compliance.