Reducing Analytical Opacity in Decision Support: An AI-Enabled Framework for Traceable and Explainable Reporting
Mingqin Yu, Felix Ter Chian Tan, Fethi Rabhi · Proceedings of the ... Annual Hawaii International Conference on System Sciences/Proceedings of the Annual Hawaii International Conference on System Sciences · 2026
Organizations increasingly rely on analytical metrics—computed indicators derived from data and models—to support ESG decision-making and disclosure. Yet these metrics are often produced through opaque workflows, making it difficult to trace their origin, recompute their values, or explain their meaning. This lack of transparency undermines auditability, adaptability, and stakeholder trust. To address this, we propose an analytical capability framework structured around three core functions: traceability, computability, and explainability. The framework integrates semantic technologies such as knowledge graphs with AI components like large language models to enhance transparency across the ESG metric lifecycle. We evaluate this framework through a positivist case study of WP, a leading financial institution in the Asia-Pacific region, using empirical disclosures to assess explanatory adequacy. Our findings show how capability gaps can be systematically diagnosed and addressed through system redesign. The proposed model supports more interpretable and accountable ESG reporting—meeting critical needs in both current practice and emerging regulatory expectations.