Automated fact verification has become indeed very important against the outgrowth of misinformation and disinformation in the digital age. This paper aims to conduct a thorough review of recent techniques using different machine learning (ML), deep learning (DL), and natural learning language processing (NLP)-based methods for fact verification problems. This paper presents an overview of the progress in automated fact-checking with particular attention to the hybrid approaches that combine several approaches to better overall performance. A major contribution of this review covers elements like feature extraction, knowledge graph integration, transformer-based architectures, i.e., BERT, and hybrid models incorporating machine learning with deep learning methods. Additionally, it highlights emerging progress in claim-checking pipelines, evidence retrieval systems, and stance classification approaches, which are essential components of scalable and effective fact verification systems. We cover hurdles such as bias in datasets, explainability, and computational complexity of hybrid models, and future avenues such as integration of multimodal data and real-time verification. The objective of this paper is to summarize the state of the art of automated fact verification, with a view to revealing opportunities for researchers, practitioners, and policymakers in this field of major importance. The review brings together current developments and approaches, thus helping to move toward robust and scalable systems to combat the global challenge of disinformation.