An Explainable Agentic AI-Based Decision Support System for Ethical and Risk-Aware Business Contract Vulnerability Detection

Business contracts frequently contain hidden financial, operational, or intellectual property risks that disproportionately affect less-informed parties. Existing AIbased contract review tools rely primarily on static text analysis, limiting their ability to identify systemic and downstream vulnerabilities. This paper presents an explainable agentic AIbased decision support system designed to autonomously analyze business contracts and detect critical vulnerability hotspots, including imbalanced indemnities, restrictive noncompete clauses, ambiguous intellectual property terms, and concealed penalties. The proposed multi-agent architecture integrates four components: a parsing agent for clause extraction and classification, a compliance agent for legal and ethical validation, a predictive risk agent for financial and reputational scenario simulation, and an explainability agent that produces transparent, human-readable insights. Evaluation on a curated set of anonymized business agreements demonstrates improved detection accuracy, reduced false positives, and enhanced interpretability compared to rule-based approaches. Results indicate that agentic AI can support ethical, transparent, and risk-aware contract governance.