This study presents a novel approach to detecting cyber attacks using machine learning, with a focus on dynamic attacks and identifying spyware through deep learning methods. Researchers are exploring machine learning as a means to enhance monitoring, as traditional protection methods often struggle to keep pace with evolving cyber threats. Our proposed method aims to address this challenge by utilizing deep learning algorithms to effectively identify dynamic attacks and categorize different types of malware. Previous research has demonstrated that machine learning can successfully detect various cyber attacks. The increasing sophistication of threats, The need for advanced detection techniques that can instantly identify multifaceted and changing attacks has increased due to the growing frequency of threats like spyware, phishing emails, even denial-of-service assaults. Our suggested method uses cutting-edge deep learning techniques, such as convolutional neural networks (also known as CNNs) and neural networks with recurrent neurons (RNNs), to address this demand by identifying indicators of malicious activity in system logs, network traffic, and user behaviors. Using large-scale datasets of known cyberthreats, our approach trains these deep learning models to identify complex patterns and behaviors that indicate different kinds of attacks. Our method also highlights how important the technique of deep learning was is to improving malware identification and reaction analyses.