<![CDATA[This study analyses the impact of artificial intelligence (AI) on bank performance and financial stability in Indonesia using a combined empirical and simulation approach. Employing monthly panel data from 50 Indonesian banks during the period 2017–2023, the empirical analysis applies Feasible Generalised Least Squares (FGLS) and staggered Difference-inDifferences (DiD) estimators following Callaway and Sant’Anna to identify the effects of AI adoption. The results show that AI adoption increases Return on Assets (ROA) by 0.116 percentage points and reduces Non-Performing Loans (NPLs) by 0.068 percentage points, while AI chatbot implementation raises ROA by 0.189 percentage points and lowers the Loan-to-Deposit Ratio (LDR) by 3.18 percentage points. In contrast, AI training is associated with a short-run decline in ROA, reflecting implementation and adjustment costs. Staggered DiD estimates indicate that bank performance weakened during the COVID-19 period due to heightened liquidity pressures but stabilised in the post-pandemic phase, thereby contributing to improved financial stability. Complementing the empirical findings, a system dynamics simulation demonstrates that AI investments initially constrain profitability growth but yield accelerating gains over time through efficiency improvements and risk mitigation. Overall, the study highlights the importance of phased AI adoption and institutional readiness in strengthening bank performance and financial stability in Indonesia.]]>
Copyrights © 2026
