Explainable Human-AI Framework for Sustainable Multi-Objective Policy Design
Document Type
Conference Proceeding
Source of Publication
2026 17th Student Research Conference on Applied Computing Src 2026
Publication Date
4-22-2026
Abstract
Artificial intelligence is increasingly applied in sustainability governance, yet many optimization systems collapse complex objective spaces into single prescriptive outputs, limiting transparency and deliberative legitimacy. This paper proposes an Explainable Human-AI Cooperative Framework for multi-objective policy design that preserves Pareto-efficient policy portfolios rather than scalarized optima. The architecture integrates multi-objective optimization, objective-level explainability, uncertainty exposure, distribution-Aware equity modeling, and iterative human value parameterization within a governance-Audited structure. Unlike conventional Explainable Artificial Intelligence (XAI), it exposes structural trade-offs among emissions, cost, equity, and robustness. A UAE national carbon allocation case study demonstrates how diverse sustainability priorities can be navigated without collapsing normative diversity, reframing AI from decision authority to transparent deliberative infrastructure.
DOI Link
ISBN
[9798319510167]
Publisher
IEEE
Disciplines
Computer Sciences
Keywords
explainable AI, fairness, human-in-The-loop, multi-objective optimization, Pareto frontier, policy design, sustainability governance, uncertainty analysis
Scopus ID
Recommended Citation
Alsuwaidi, Mahra; Almansoori, Meera; and Abugabah, Ahed, "Explainable Human-AI Framework for Sustainable Multi-Objective Policy Design" (2026). All Works. 8218.
https://zuscholars.zu.ac.ae/works/8218
Indexed in Scopus
yes
Open Access
no