2024年从银行业的人工智能中提取价值：重新连接企业报告

Investment Rating - The report emphasizes the need for banks to transition from experimentation to becoming AI-first institutions to unlock material financial value from AI technologies [1][3]. Core Insights - AI is transforming the banking industry, but many banks remain in the experimental phase. To thrive, they must adopt AI technologies enterprise-wide [3][18]. - Successful AI transformation requires investment across multiple interdependent layers of the organization, as underinvestment in any one area can hinder overall progress [4][22]. - Leading banks view AI as a transformational tool, focusing on core strategic priorities such as revenue enhancement and customer satisfaction [22][68]. Summary by Sections AI Transformation Essentials - Banks need to modernize core technology, reimagine customer experiences, and enhance decision-making processes through AI [4][18]. - A comprehensive AI capability stack is essential, including automated cloud provisioning, streamlined architecture, and a central AI control tower [4][54]. Business Areas for AI Transformation - Banks can identify high-impact subdomains for AI transformation, such as customer underwriting, risk management, and sales and marketing [7][12]. - The report identifies that a typical bank has around 25 subdomains that could benefit from AI rewiring, with a focus on those with high business impact and technical feasibility [24][27]. Multiagent Systems - Multiagent systems are highlighted as a key component in automating complex banking workflows, improving productivity by 20 to 60 percent in tasks like credit memo preparation [21][45]. - These systems can act as virtual coworkers, capable of planning and executing tasks, thus enhancing both customer and employee experiences [61][67]. Investment in AI Capabilities - To extract value from AI, banks must invest across the entire AI capability stack, including enterprise data, machine learning operations, and core technology [68][77]. - The report stresses the importance of a full-stack approach that combines generative AI with traditional analytics and digital tools [72][78].