AI-Powered Cloud Security: Leveraging Large Language Models for Threat Detection and Risk Mitigation

Abstract

As cloud adoption grows, so do the risks associated with cyber threats, data breaches, and compliance violations. This paper introduces a Large Language Model (LLM)-based cybersecurity framework for cloud environments, integrating natural language processing (NLP) with threat intelligence and anomaly detection. The proposed system automatically analyzes security logs, detects malicious patterns, and generates real-time incident response recommendations. Experimental evaluation shows that LLM-driven security systems enhance detection accuracy by 40% compared to traditional cloud security mechanisms, paving the way for AI-augmented, self-healing cloud infrastructures.