The loss of crucial and sensitive data leaves a harmful impact on the user’s profile. It is indispensable to protect the sensitive data of the user, against the network attack. It is very difficult to detect unwanted and harmful packets that circulate over the networks because of variations in data and information. IT Services in the business domain enable users to be involved in a wide range of activities related to manager operations, technical expertise, network infrastructure management, and information technology consultations. These services are vulnerable to different kinds of attacks due to their weakness and flaws that open the door for network breaches and weaknesses. Artificial Intelligence replicates the rational capabilities of the human brain and emerges as a smart enhancement to different fields such as education, hospitals, manufacturing, smart homes, and much more. A Generative Adversarial Network (GAN) consists of two models that is Generator and a Discriminator. The Generator produces the fake samples and the Network Intrusion Detection systems are trained on that data. GAN provides a complete framework for approaching Generative Artificial Intelligence. The framework is trained with a collaborative Machine Learning Model K-nearest neighbor and Random Forest and with the help of different matrices the model is evaluated for performance. This system uses the GAN to generate synthetic data that is helpful for the classification of novel attacks and increased accuracy. A dataset IoTID20 is used for experimentation that is generated in a real Internet of Things (IoT environment). This system helps secure IT Services in an IoT environment. The result shows that the K-nearest neighbor has the better performance as it achieves an accuracy of 97.51% as compared to the Random Forest which achieves an accuracy of 96.20%.