Workflow
智能授信
icon
Search documents
【招银研究|行业深度】数字金融之AI+银行——大模型与银行数字化转型的三组关系
招商银行研究· 2025-07-18 09:00
Core Viewpoint - The development of large models and AI technologies is creating new paths for technological empowerment in the banking industry, aiming to enhance asset organization efficiency and reduce operational costs through digital transformation [1]. Group 1: Relationship between Large Model Capabilities and Banking Application Scenarios - Large model technologies have achieved significant breakthroughs in natural language processing, including content generation, information extraction, and dialogue interaction, which align well with the knowledge-intensive characteristics of the banking industry [1][9]. - Applications in the front office include knowledge bases and intelligent customer service, with examples showing a 10% reduction in call duration and an 80% decrease in labor costs [5][14]. - In the middle office, intelligent credit assessment has reduced due diligence report writing time from one week to five minutes, indicating a potential shift towards real-time, comprehensive, algorithm-intensive credit review processes [1][21]. - The backend development has seen improvements in code generation efficiency, with several banks reporting a 20%-30% increase in productivity [5][24]. Group 2: Generative AI vs. Discriminative AI - Generative AI excels in creating new content from unstructured data but faces challenges such as high computational costs and poor interpretability, while discriminative AI (e.g., logistic regression, decision trees) is widely used in banking risk control due to its efficiency and accuracy [2][31]. - Future collaboration between generative and discriminative AI is expected to create two models: a "hub-and-spoke model" where generative AI disassembles tasks and integrates results, and a "serial model" where both types work at the same level [2][39]. Group 3: AI and Banking Digital Transformation - The application of large models aims to drive digital transformation in banks, which requires deep changes in business processes supported by strategic planning, organizational collaboration, and technology implementation [3][7]. - Historical analysis shows that significant technological innovations in the banking sector have always been accompanied by process adjustments, emphasizing the need for comprehensive transformation commitment from financial institutions [3][56]. - The digital transformation success rate in enterprises is only 16%, highlighting the importance of integrating digital technology deeply into business processes for sustained competitive advantage [51][55].