Demands for flexible and seamless system management necessitated the growth of software-defined networks (SDN). Yet, securing these environments with effective measures is critical as SDN continues to evolve into more intricate architectures. Intrusion detection is paramount among these measures; thus far, studies suggest that artificial intelligence (AI) approaches may be helpful in this domain. By systematically examining relevant works addressing AI-based intrusion prevention strategies within hyper-evolved SDN settings, our review aims to present an inclusive evaluation alongside suggesting areas requiring additional scrutiny. This research introduces readers to key concepts related to SDN and how deep learning algorithms, machine learning algorithms, and neural networks can be applied for effective intrusion detection within an SDN environment. Drawing from existing literature on this subject matter, our analysis critically examines the benefits and drawbacks of these AI-based techniques while highlighting gaps in knowledge requiring further research attention. Some areas include real-time protection capabilities, scalability concerns, and seamless integration with different security mechanisms. We then present future research directions in this area. This literature review employs a systematic approach to elucidate the current research on using AI methods to detect intrusions in SDN.