The rapid dissemination of information in the age of social media has completely changed how people communicate and keep informed. But the same platforms that make it possible to communicate instantly have also made it easier for rumor and incorrect information to spread. Unverified claims that often proliferate without proof, rumor can have negative effects on the economy, politics, and society. Social media platforms like Facebook, Reddit and twitter disseminate a lot of information, this study explores the use of machine learning and Explainable AI (XAI) approaches for rumor identification on social media data. The speed and volume of data created by social media make it extremely challenging to find patterns for detecting rumor manually. Deep learning networks like Long Short-Term Memory (LSTM), Recurrent Neural Networks (RNNs), and Convolutional Neural Networks (CNNs) and a variety of machine learning techniques, including Support Vector Machines (SVMs), can automatically recognize patterns and categorize social media data. The objective of this review is to compare multiple studies on rumor detection integration with machine learning and XAI, their research gaps and future scope.