The rapid expansion of the digital landscape has made cybersecurity a top priority for both individuals and organizations. The increased frequency of cyberattacks has evolved drastically over the last few decades to keep up with ever-changing technologies. This paper reviews how Artificial Intelligence (AI) technologies can address cyberattacks with emphasis on the key AI technologies, their performance, and the potential implementation restraints. Articles were obtained from the following reliable academic databases: IEEE Xplore, Research Gate, and Elsevier. Results from the review showed that traditional intrusion detection systems, which are signature-based, fail to detect cyberattacks in real-time. Machine Learning (ML) and Deep Learning (DL) Models are being used as a potential solution for cyberattack detection in real-time. Additionally, the findings indicate that AI can support automatic vulnerability scanning, using techniques such as data mining and expert systems that can greatly decrease the time of detection and the cost of operation. The issues of interpretability and the issues of scale related to the application of AI solutions to large-scale systems are some of the AI challenges that were identified in the review. The findings outline the potential and the existing shortcomings of AI in cybersecurity, which require further research to enhance explainability, performance, and scalability of future AI-based security systems.