Examining the Barriers to Implementing AI-Based Bankruptcy Prediction Models

Abstract

This study aims to identify the barriers to implementing AI-based bankruptcy prediction models in Iranian organizations. This qualitative study was conducted using an inductive content analysis approach. Data were collected through semi-structured interviews with 30 financial managers, AI specialists, and economic analysts in organizations based in Tehran. Participants were selected purposefully, and data collection continued until theoretical saturation was reached. The data analysis was performed using NVivo software, and the coding process consisted of three stages: open coding, axial coding, and selective coding. The results indicated that barriers to implementing AI-based bankruptcy prediction models fall into four main categories: technological barriers, organizational and managerial barriers, financial and economic challenges, and human and cultural factors. Key obstacles identified included infrastructure deficiencies, data analysis complexity, managerial resistance to change, high implementation costs, and a lack of trust among employees in AI-driven decisions. The findings suggest that for the successful adoption and implementation of AI-based bankruptcy prediction models, companies should focus on both technological infrastructure development and training programs for managers and employees. Additionally, policy support and financial incentives from governmental and regulatory bodies can help mitigate existing barriers.