Authors:
Abhilash Reddy Pabbath Reddy
Addresses:
Department of Information Technology, Axle Info, Cumming, Georgia, United States of America.
Abstract:
The widespread use of multi-cloud infrastructure has redefined business IT in the form of increased redundancy, scalability, and flexibility. The increased complexity also introduces new security risks, particularly in detecting and mitigating advanced threats across diverse platforms. This paper conducts in-depth research on AI-driven anomaly detection systems for enhanced cybersecurity in multi-cloud networks. Using machine learning and deep learning methods, such as autoencoders and clustering algorithms, the system autonomously detects threat behaviours and anomalies. Supervised and unsupervised models are combined to create dynamic baselines for detection. Real-time auditing of traffic, log correlation, and anomaly score computation on emulated multi-cloud harvested data are characteristics of the approach. The data are computed with AI models and inspected for the detection of false positives, speed, and latency. An AI-driven platform responds much quicker than traditional SIEM solutions and is also more precise. Anomaly trends are displayed through contour and waterfall plots, while performance measurements are recorded against comparison tables. The architecture diagram specifies the structure of the data ingestion layer, AI engines, and decision endpoints. Experiments confirm that the system effectively identifies weak threats across multiple cloud platforms. Performance enhancement, pattern recognition, and comparison with current models are addressed. Future work and limitations are the application of federated learning and adaptive algorithms for improved detection in distributed networks.
Keywords: Multi-Cloud Security; Anomaly Detection; Artificial Intelligence; Cyber Threats; Machine Learning; Distributed Networks; Cloud Platforms; Comparison Tables; Cloud Infrastructures; Cybersecurity Management.
Received: 07/09/2024, Revised: 28/11/2024, Accepted: 10/01/2025, Published: 07/06/2025
DOI: 10.64091/ATICS.2025.000132
AVE Trends in Intelligent Computing Systems, 2025 Vol. 2 No. 2 , Pages: 77-86