Policy-Aware Secure Data Governance in Distributed Information Systems Using Explainable AI Models

Abstract

In the modern digital age, distributed information systems have become the key infrastructure in terms of storing and sharing data as well as processing. These systems are cross-administrative and pose considerable problems to the data government that is secure, hence complies with both the organizational and regulatory policies. Such systems are decentralized, which is why these systems create additional complications associated with data privacy, security, trust, and compliance. To overcome these difficulties, we plan to introduce a multifaceted solution to policy-aware secure data governance, leveraging Explainable Artificial Intelligence (XAI) models. Our solution is based on combining dynamic policy enforcement approaches with security rules and XAI methods aimed at improving the transparency, responsibility, and explainability of security decisions. The proposed system allows organizations to know how to maintain governance over data in distributed environments securely, and gives humans a reasonable explanation of the policy violation and access control decisions. The architecture proposed in the paper comprises the following layers involved in policy definition, policy enforcement, data auditing, and the interpretation of decisions based on XAI. We develop and deploy an explainable decision engine to production, built on SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-Agnostic Explanations), to demonstrate how the privilege of accessing particular data is granted or denied. Besides, through it, our methodology uses federated learning and blockchain to preserve the integrity and provenance of the data over decentralized nodes. An extensive literature review is carried out to show the available gaps in secure data governance and explainable AI. A well-developed experimental setup and a miscellany of case studies show the efficiency of our method to enhance policy compliance, reduce illegal attempts to enter the system, and instil trust in stakeholders. In our findings, the governance process is becoming more rigorous and reputable in detecting policy violations since it has a high level of interpretability