The outstanding development of distributed computing and the boundless reception of the Internet of Things (IoT) have acquired critical progressions different areas, including medical care, money, and shrewd urban communities. In any case, this quick extension has likewise presented new security challenges, especially as far as protecting cloud frameworks from noxious interruptions. Customary intrusion detection systems (IDS) frequently miss the mark in tending to the intricacies of cloud conditions, particularly while managing the different and dynamic information produced by IoT gadgets. To address these difficulties, this paper proposes a deep learning-based system for intrusion detection in cloud conditions. The proposed structure incorporates a mix of convolutional neural networks (CNN) and long short-term memory (LSTM) organizations to successfully dissect both spatial and transient elements of organization traffic information. This double methodology empowers the model to precisely distinguish likely dangers and decrease misleading up-sides, guaranteeing the security and unwavering quality of cloud-facilitated administrations. Exploratory outcomes exhibit the system’s prevalence over existing techniques, with an exactness of 98.2%, a discovery pace of 97.4%, and a misleading positive pace of simply 1.1%. The discoveries recommend that the proposed structure is a powerful answer for improving cloud security, especially with regards to IoT-coordinated frameworks.