The integration of artificial intelligence (AI) in oncology has revolutionized oncology by enabling precise diagnostics, personalized treatment, and improved patient outcomes with use of AI techniques such as machine learning (ML), deep learning (DL), reinforcement learning (RL), and natural language processing (NLP). This paper provides a comprehensive analysis of AI applications in oncology, evaluating state-of-the-art methodologies, their performance metrics, associated challenges, and outlining future directions for AI-driven precision oncology. Further, the overview of advanced AI technologies in oncology, alongside the workflow, highlights the several critical steps that involve integration of AI in oncology. A comparative performance evaluation of AI models highlights their effectiveness in oncology, and certain observations are drawn from this comparison such that AI-driven models, specifically convolutional neural networks, have demonstrated high accuracy in tumor detection using medical imaging, while NLP enables the extraction of critical clinical data, and deep learning-based models often achieve diagnostic accuracy exceeding 90%, yet certain challenges such as model interpretability, data scarcity, regulatory constraints, and ethical considerations hinder widespread adoption. The lack of diverse, high-quality datasets and concerns regarding AI transparency raise issues of clinical reliability and integration. The study proposes future directions to increase clinical adoption by using multimodal data fusion techniques, developing real-time applications with edge AI, advancing explainable AI, and developing standardized regulatory framework. The study concludes with insights into the most impactful AI techniques in oncology which contributes to the ongoing effort to integrate AI seamlessly into clinical oncology, ultimately advancing precision medicine and improving global cancer care.