Detecting Device-Aware Phishing Attacks in Cloud Environment Using FNN and RNN

This paper focuses on an innovative detection technique of device-aware phishing attacks on the cloud platform based on FNN and RNN. Over time, the phishing attack, especially targeting the cloud, has developed sophisticated characteristics based on specific characteristics such as IP address, geolocation and configurations of devices. The new threat model associated with cloud attacks will challenge its security. Current traditional phishing detection mechanisms, based on signature and heuristic techniques, do not learn with dynamic device-aware threats. In this sense, the developed model combines FNN and RNN capabilities. It can manage both complex data and sequential patterns of the model. The system is tested with multiple success indicators: the F1-score, knowledge, efficiency and exactness, it was evaluated with a significantly higher detection accuracy for phishing attacks. The accuracy achieved was 99%, and the precision achieved was 97% with a recall of 96%. On comparison with some of the familiar old AI models, such as RF, KNN & LR, the proposed model FNN–RNN achieved a higher level of accuracy, making it an efficient solution to the problem. The results indicate that the system enhances phishing detection in dynamic cloud environments while also improving on security, scalability and resilience against evolving threats. The approach given here presents a powerful solution for advanced phishing attacks in the cloud environment as an existing potential of deep learning models in cybersecurity.