AI Governance Risk Tiering for Sustainable Digital Infrastructure: A Systematic Review of Cybersecurity Frameworks

The rapid adoption of artificial intelligence (AI) across public services and critical infrastructure is reshaping digital governance. While AI promises efficiency and innovation, its reliance on large, high-dimensional datasets introduces privacy, bias, transparency and accountability risks that existing frameworks struggle to address. This study evaluates the maturity of current AI governance frameworks and develops an integrated risk-tiering model that connects ethical principles to auditable technical controls, aligning with Sustainable Development Goal 9 on industry, innovation and infrastructure. A systematic literature review of 450 records from major databases was conducted using PRISMA 2020 guidelines; 95 high-quality studies were analyzed using principal component analysis and k-means clustering. The analysis produced a heat map of governance frameworks, a co-occurrence network of themes, a cluster analysis of framework coverage and an integrated governance risk framework supported by a risk-tiering matrix. Findings reveal a fragmented landscape dominated by ethics/privacy-centric and compliance/risk-focused approaches, with few integrated frameworks and evident tension between privacy and security. This synthesis bridges the gap between values and practice, offering a policy-ready model for secure and sustainable AI governance.