Banking based on artificial intelligence (AI Banking) is a new phenomenon that stresses the integration of advanced AI technologies and techniques in the banking industry. The applications of AI in this industry encompass a wide array of solutions and innovations that are meant to upgrade different aspects of banking operations, customer experience, and financial services. Given the significance of incorporating AI in the banking system, the present study scrutinizes the features of AI banking and explores different challenges facing its implementation. Accordingly, the challenges and their evaluation criteria are ranked using the best-worst method (BWM), and the efficiency of the proposed solutions to address these challenges is assessed using the data envelopment analysis (DEA) method. Following this, a bi-objective mathematical model is proposed to minimize the cost of implementing the solutions and maximize their effectiveness, leading to the selection of the most suitable solution for addressing each challenge. Subsequently, the proposed model is solved using the augmented ε- constraint method, and the Pareto solutions derived from it are ranked using the DEA to identify the most efficient ones. Consequently, the most effective solutions for overcoming the challenges are identified and recommended. The findings indicated that the primary challenge in AI banking, namely the reducing the quality and availability of data, can be effectively addressed by developing and implementing data governance policies.