Ransomware has emerged as one of the most critical cybersecurity threats, targeting individuals, enterprises, and critical infrastructure worldwide. The rapid evolution of ransomware techniques [1] [2], including Ransomware-as-a-Service (RaaS), double extortion, and fileless attacks, has rendered traditional signature- based security mechanisms increasingly ineffective [3] [4] [5]. This paper presents a comprehensive analytical study of modern ransomware attacks, examining their evolution, attack vectors, and real-world incidents such as WannaCry, Ryuk, and the Colonial Pipeline attack [6] [7]. A systematic literature review is conducted to evaluate existing detection and mitigation strategies, with a particular focus on artificial intelligence-driven techniques, including behavioural analysis, anomaly detection, and machine-learning-based threat intelligence. The effectiveness of multi-layered defence mechanisms, including endpoint security, blockchain- based data integrity, and Zero Trust architectures, is critically assessed. The findings highlight that AI-based detection models significantly improve early identification of ransomware, while proactive threat intelligence and coordinated defence frameworks enhance organisational resilience. This study contributes practical insights for strengthening ransomware defence strategies and outlines future research directions for adaptive and intelligent cybersecurity systems.