Abstract

An authentication framework is a systematic technique for authenticating the identification of persons or organizations that use a system or service. The Multi-Device continuous authentication framework improves security by dynamically authenticating users on multiple devices. This research aims to develop an intelligent Artificial intelligence (AI)-based multi-device continuous authentication framework for enhancing cloud security in financial sectors. We established an innovative architecture named Secure Sync (SS) continuous authentication system for privacy-preserving and providing continuous authentication. We obtained data from smart office infrastructure, which includes two users’ several days of usage of cell phones, tablets, laptops, and desktop computers. Adam-boosted long short-term memory (ABLSTM) approach is proposed to predict the authenticate users depending on their behaviors in Secure Sync (SS) architecture. Our proposed model enhances the constraints of single-device approaches by incorporating further contextual data from varied devices. The result analysis phase evaluates the classification efficacy of the proposed model using various metrics such as precision – 99.37%, recalls –99.4%, F1 score – 99.42% and FPR- 0.0018%. We also conducted a comparison analysis with other comparable existing methods. The experimental results demonstrate that the proposed model performed better than other conventional methods for enhancing cloud security through multi-device continuous authentication in financial sectors.