Agentic AI in Airline Management: A KPI-Governed Architecture for Trust-Based Autonomy, Strategic Co-Leadership, and Operational Excellence

Abstract

This study investigates the strategic integration of agentic artificial intelligence (AI) into airline management systems using a KPI-governed architectural model based on the Perception–Cognition–Strategy–Action (P–C–S–A) framework. The research aims to address the lack of standardized, explainable, and ethically governed AI frameworks in aviation by proposing a multi-layered model that enhances real-time perception, predictive cognition, strategic alignment, and autonomous action. Employing a qualitative, systematic literature review of over 1000 scholarly sources published between 2016 and 2025, the study analyzes emerging tools such as IoT-driven perception systems, XAI technologies (e.g., SHAP, LIME), simulation platforms (e.g., AnyLogic, Simio), and digital twins. Findings reveal that embedding KPI-linked layers significantly improves situational awareness, operational transparency, and strategic co-leadership between human managers and AI agents. The research further identifies critical KPI architectures Balanced Scorecard, ESG-aligned metrics, and CASK indicators as foundational to trustworthy AI orchestration. The study offers actionable recommendations for practitioners and policymakers, including implementation of ESG-compliant automation protocols, transparent decision workflows, and ethics-governed RPA integration. The results contribute to both theoretical models of digital transformation and practical strategies for certifiable AI deployment in airline ecosystems.