Detecting Identity Theft Attacks in Cloud Security based on Artificial Intelligence Techniques

Kadhim Q. K., Ahmed S. T., Kadum J., Turki E. M., Kadhim A. K.

14th International Conference on Applied Innovations in IT, ICAIIT 2026, Kothen, Germany, 31 March 2026, vol.14, pp.1093-1099, (Full Text)

  • Publication Type: Conference Paper / Full Text
  • Volume: 14
  • City: Kothen
  • Country: Germany
  • Page Numbers: pp.1093-1099
  • Keywords: Denoising AutoEncoder (DAE), Detecting Identity, Long Short-Term Memory (LSTM), Theft Attacks
  • Istanbul Gelisim University Affiliated: Yes

Abstract

Identity theft attacks are among the most critical security challenges in cloud computing environments, as they allow malicious actors to gain unauthorized access to sensitive data and cloud-based services. With the rapid expansion of cloud computing applications, the need for intelligent and proactive defense mechanisms has become increasingly vital. This study introduces an Artificial Intelligence (AI) framework designed to detect and mitigate identity theft attempts by leveraging the Denoising AutoEncoder (DAE) and Long Short-Term Memory (LSTM) algorithms. The DAE component efficiently removes noise and extracts essential features from input data, while the LSTM network captures temporal dependencies to enhance anomaly detection. The proposed model was evaluated using a conventional cloud infrastructure, achieving a high detection accuracy of 94.90% with a notably low false positive rate. These results highlight that integrating AI-driven models such as DAE and LSTM can substantially strengthen cloud computing security by enabling early detection and prevention of identity theft attacks.