https://jjcit.org/paper/209 10.5455/jjcit.71-1694088480 Sanaa Ali Jabber,Soukaena H . H ashem,Shatha H. Jafer Cyber threats,Intrusion detection,Cloud environments,Long short-term memory (LSTM),Snake algorithm,Intrusion detection systems (IDSs) 1 4884 1182 7-Sep.-2023 29-Oct.-2023 11-Nov.-2023 In the epoch of digital transformation, cloud computing remains paramount, acting as the linchpin for a plethora of services from enterprise solutions to day-to-day consumer applications. Yet, its expansive nature has invariably rendered it susceptible to a myriad of cyber threats, necessitating advanced, adaptive defense mechanisms. This paper introduces a novel intrusion-detection method tailored for cloud environments, ingeniously amalgamating the temporal pattern-recognition capabilities of Long Short-Term Memory (LSTM) networks with the heuristic finesse of the Snake algorithm. Our research meticulously delineates the LSTM-Snake model’s design, implementation and exhaustive benchmarking against prevailing approaches for a rigorous and comprehensive evaluation of cloud-based intrusion-detection systems and by using the TON-IOT dataset, a carefully curated dataset tailored for cloud-centric applications. The experimental results underscore the model’s prowess, registering a commendable 99% accuracy rate in intrusion detection; a marked improvement over current state-of-the-art methodologies. The ensuing discussions offer insights into the model’s practical implications and potential limitations.