This paper presents a comprehensive survey of automatic text summarization techniques, thoroughly exploring the methodologies, datasets, and evaluation metrics that shape the current field. The study covers both extractive and abstractive summarization approaches, detailing their strengths, limitations, and applications. Through a detailed literature review, it examines the evolution of these techniques, contrasting traditional methods with recent advancements in machine learning and deep learning. A systematic categorization of existing methods is provided, alongside an analysis of commonly used datasets, highlighting their impact on model performance. The effectiveness of different summarization techniques is evaluated using metrics like ROUGE, BLEU, and BERTScore, while also addressing the limitations of these metrics. Emerging challenges, such as maintaining factual accuracy in abstractive summaries and coherence in extractive ones, are discussed, along with the technical and ethical issues in scaling these techniques for real-world applications. The paper concludes by identifying key areas for future research, including the development of hybrid models, more diverse datasets, and refined evaluation metrics. This survey offers valuable insights for researchers and practitioners, making it a critical resource for understanding the current trends and future potential of automatic text summarization.