The occurrence of harmful software, also known as malware is on the rise with certain types of malware becoming adept at camouflaging themselves within a system through sophisticated techniques. It is crucial to detect malware early on to prevent widespread damage to computer systems and the Internet. Numerous techniques for identifying malware have been created in recent times. Despite this, detecting malware remains a difficult task so this paper proposes the utilization of MobileNet-based CNN classification for malware detection. The initial dataset undergoes pre-processing to simplify the subsequent processes. Data preprocessing is a crucial step in data preparation where raw data undergoes various processing techniques to make it suitable for further processing. Data visualization involves transforming data into visual representations, such as charts or diagrams, to enhance comprehension and extract valuable insights for humans. These visualizations aid in understanding the data structure and detecting any anomalies present. MobileNet is specifically designed to reduce the parameter count, enhance training speed and provide accurate predictions. MobileNet is a convolutional neural network (CNN) design created specifically for the tasks of image classification and mobile vision applications. This makes it ideal for running on mobile devices or for implementing transfer learning techniques. To evaluate the model’s performance, the confusion matrix is generated and this technology is well-suited for mobile devices, embedded systems and low-power computers that maintain a high accuracy without sacrificing computational efficiency. This work is implemented in python software.