AI and Machine Learning Advances

The high rate of Cloud computing and multi-tenant architectures adoption has greatly complicated Identity and Access Management (IAM) systems security. Conventional rule-based access control systems can be very difficult to detect misconfigurations and threats that emerge in dynamic clouds. The study provides an artificial intelligence-based discussion of an IAM security setup to establish the key attributes that affect access control security in multi-tenant cloud environments. The dataset that consisted of 100,000 IAMs with 50 parameters that reflect the security level was examined to determine the predictability of the machine learning in predicting the security strength. The data set contains authentication mechanisms, authorisation models, governance frameworks and network security control with a security score as the target variable. To model the link between the IAM parameters and the security scores, a Random Forest classifier was used. The experimental findings indicate that the suggested method works with an accuracy of 82.6 per cent in forecasting IAM security settings. In the analysis of the feature importance, it is possible to identify user identity management, access levels, data governance frameworks, geolocation restrictions, and access control by tokens as some of the most powerful security factors. The results indicate the relevance of tiered security features that entail a combination of identity checks, policy controls, and network defence. The study can be used to design AI-based threat modelling and mitigation strategies that enhance IAM security in multi-tenant cloud system applications.