余滨表示："AI已从'试试看'走向'必须做'，正从成本中心转变为服务、营销与价值创造的核心。"目 前，蚂蚁数科为银行提供的AI应用已从早期的智能客服、知识问答等单点场景，扩展至覆盖营销、风 控、投顾、理赔等核心业务的全栈智能体系统。据试点银行反馈，借助理财经理数字分身，人均服务客 户数从200人提升至2000人，有效覆盖了原本难以触达的中长尾客户，带动收入增长约20%。 10月30日，在2025金融街论坛上，蚂蚁数科AI业务总裁、蚂蚁集团副总裁余滨发表演讲时透露，当前 不少城商行正积极拥抱金融大模型与智能体技术，寻求业绩突围。蚂蚁数科致力于支持银行构建自主可 控的金融大模型，江浙沪地区的部分城商行在员工工作效率、营销转化率、营收规模等方面取得显著增 长。 伴随净息差收窄、营收增速放缓以及同质化竞争加剧，城商行面临前所未有的挑战。原本的区域优势减 弱，大型银行持续挤压其生存空间。在此背景下，越来越多的银行将AI视为"弯道超车"的关键路径。 浙江一家头部城商行则与蚂蚁数科合作，系统性构建了"算力-平台-模型-应用"全栈AI能力，落地30多个 智能体场景，覆盖客户服务与内部运营两端，为未来十年的AI转型奠定坚实基础 ...

北京商报讯(记者刘四红)10月30日，在2025金融街论坛上，蚂蚁数科AI业务总裁、蚂蚁集团副总裁余滨 发表演讲时透露，当前不少城商行正积极拥抱金融大模型与智能体技术，寻求业绩突围。蚂蚁数科致力 于支持银行构建自主可控的金融大模型，江浙沪地区的部分城商行在员工工作效率、营销转化率、营收 规模等方面取得显著增长。 据了解，蚂蚁数科的金融数智化服务已覆盖100%的国有股份制银行、超60%的地方性商业银行及数百 家金融机构。 以上海银行（601229）为例，蚂蚁数科助力其打造的AI手机银行，以"对话即服务"为核心，用户通过自 然语言交互即可办理转账、还款、理财咨询、养老金查询等十余项高频业务。用户不用在复杂菜单中翻 找功能入口，操作门槛低了，老年客户满意度更高了，业务转化率提升了10%。 浙江一家头部城商行则与蚂蚁数科合作，系统性构建了"算力-平台-模型-应用"全栈AI能力，落地30多个 智能体场景，覆盖客户服务与内部运营两端。 余滨强调，不同金融机构的资源禀赋与发展阶段各异，应选择适合自身的AI演进路径。预算有限的区 域性银行，可以选择先从场景应用入手，按效果付费，逐步加大投入；也有很多银行优先将原有的手机 Ap ...

Core Insights - Ant Group's AI business is actively supporting city commercial banks in adopting financial large models and intelligent agent technologies to enhance performance amidst challenges like narrowing net interest margins and intensified competition [1][2] - The shift from AI being a trial to a necessity is emphasized, with AI transforming from a cost center to a core element of service, marketing, and value creation [1] Group 1: AI Adoption in Banking - Many city commercial banks are leveraging AI as a key strategy for performance improvement, with significant growth in employee efficiency, marketing conversion rates, and revenue scale reported in regions like Jiangsu, Zhejiang, and Shanghai [1] - The implementation of AI applications has expanded from basic functions like intelligent customer service to comprehensive systems covering marketing, risk control, investment advisory, and claims processing [1] Group 2: Case Studies and Results - For instance, Shanghai Bank's AI mobile banking service allows users to perform over ten high-frequency tasks through natural language interaction, resulting in a 10% increase in business conversion rates and higher satisfaction among elderly customers [2] - A leading city commercial bank in Zhejiang has developed a full-stack AI capability with over 30 intelligent agent scenarios, enhancing both customer service and internal operations [2] Group 3: Strategic Approaches to AI Implementation - Different financial institutions are advised to choose AI evolution paths that suit their resources and development stages, with options ranging from starting with application scenarios to upgrading existing mobile apps into AI-driven platforms [2] - Some banks are establishing dedicated teams to create comprehensive knowledge bases and datasets, utilizing Ant Group's financial reasoning models to build autonomous "financial brains" for end-to-end business process reengineering [2] Group 4: Market Coverage - Ant Group's financial digitalization services have reached 100% of state-owned joint-stock banks, over 60% of local commercial banks, and hundreds of financial institutions [2]