Challenges on Artificial Expert Acceptance in AHP Analysis ^*

Abstract

Mirroring human specialists with artificial intelligence-based expertise presents new opportunities and threats for academic research and development. Despite their potential, the integration of artificial experts in multi-criteria decision-making remains limited, particularly in Analytic Hierarchy Process (AHP) applications. The goal of this study is to explore the artificial expert adoption challenges in AHP analysis and present the practicality of artificial experts in the analysis. To identify the challenges, a systematic literature review (SLR) was conducted. The results of the SLR presented 26 challenges that were categorized into five categories: ethical, legal, financial, technological, and societal. An AHP model was constructed based on the framework and evaluated using artificial expert judgments generated via a generative artificial intelligence (AI) tool (ChatGPT-4o). Findings reveal that ethical concerns, such as bias and misuse of technology, are the most critical barriers.The main contribution of this study consists of the development of a structured framework of adoption challenges, demonstration of generative AI for expert judgment creation in AHP, and prioritization of challenges using an AI-assisted decision model. This study presents a novel proof-of-concept demonstrating the feasibility of using generative AI to simulate expert input within AHP analysis, while providing practical insights to support responsible and effective AI adoption in structured decision-making.

Ähnliche Arbeiten