Cloud computing offers flexibility and scalability but remains vulnerable to threats such as data theft, unauthorized access, and privacy breaches. To address these challenges, we propose a Mamdani fuzzy inference system–based framework that dynamically adapts security measures to evolving conditions. The model integrates multi-factor authentication data, behavioral patterns, and anomaly-rich datasets to improve resilience against phishing, replay, and brute-force attacks. Experimental evaluation demonstrates significant improvements in detection accuracy (up to 92%), reduction in false positives (by 15%), and faster recovery time compared to baseline models. Visual analyses and comparative results validate the robustness of the approach, while the system’s modular design enables seamless integration into diverse cloud environments. This work not only strengthens authentication and privacy protection but also highlights the potential of fuzzy logic to build adaptive, secure, and intelligent cloud infrastructures.