As the internet has flourished, the connectivity between devices has increased. This has propelled the Internet of Things technology to reach new heights. But sadly, with these advancements, cyber-attacks have also increased. Consequently, these attacks have posed significant threats to every potential IoT technology user and IoT device. Such malpractices can result in substantial losses of capital and intellectual property alike. The need for developing a robust system for detecting Cyber Security threats has become a crucial operation to prevent such losses. The proposed paper establishes Deep-Learning techniques to help detect malicious attacks on an IoT architecture and prevent unwanted intrusion we've described a single strategy for identifying stolen data from software and malware throughout the Internet of Things (IoT) network in this post. The TensorFlow platform is a prime candidate to develop Deep Learning algorithms to classify stolen programming with source code literary theft. Google Code Jam (GCJ) is an international programming competition administered by google. GCJ collects data annually to examine the true nature of theft of utilizations. The use of the Mailing Dataset is prevalent to gather the malware samples. The use of Deep Learning techniques to detect Cyber Security threats presents a novel yet efficient approach to solving practical issues and shows promise for the future.