In an attempt to improve efficiency and care delivery, the healthcare industry has adopted the digitization of patient data through Electronic Health Records (EHRs). However, due to this digital revolution, EHR systems are becoming more appealing targets for ransomware attacks, which jeopardize patient safety, interrupt clinical operations, and jeopardize data integrity. The purpose of this review is to explore the AI-driven cybersecurity frameworks to prevent ransomware attacks on EHRs. The findings indicate that AI technologies improve cybersecurity by enabling proactive defense tactics, lowering false positives, and offering real-time threat detection. While deep learning models enhance the detection of complex infections, machine learning and behavioural analytics are excellent at spotting unusual activity. Although issues like data privacy concerns, resource needs, and the demand for ongoing system training still exist, integrating AI with conventional security measures may further increase organizational resilience.