A vital organ located in the top right corner of the belly; the liver is responsible for regulating the body’s general functioning. Waste removal, nutrition digestion, hormone regulation, and toxic substance detoxification are just a few of its many important functions. Failure to detect liver dysfunction in a timely manner can result in serious health complications or even death. Liver dysfunction can be caused by various diseases and conditions, including fatty liver, hepatitis, fibrosis, cirrhosis, drug-induced liver injury, and hereditary disorders. New developments in ML and DL have the potential to aid doctors in their work by revealing intricate patterns and risk factors that conventional diagnostic tools could miss. Better patient outcomes may be possible because of these technologies’ ability to diagnose liver disorders earlier and with better precision. Unfortunately, clinical adoption is hindered by the opaqueness and lack of transparency in decision-making caused by many high-performing ML/DL models, sometimes known as the “black box” problem. This paper primarily aims to offer a thorough examination of current methods used to predict and diagnose liver disease using Explainable AI (XAI), ML, and DL. By providing clear explanations of the reasoning behind the model’s predictions, interpretable AI models can help close the confidence gap in clinical settings. This review seeks to assist researchers and practitioners in creating AI systems for liver disease management that are more dependable, explicable, and clinically useful by examining present approaches, obstacles, and future directions.