值得关注的是，以智能体为代表的应用类项目的中标数量，正加速赶超算力类项目，逐步占据招投标市场的更大份额，此前以算力采购为主导的市场格局正 在悄然转变。银行业领跑大模型采购热潮 2025年，银行业正领跑大模型采购热潮，无论从采购数量、金额还是类别上看，均呈现显著增长态势。 据长期关注大模型市场中标情况的"智能超参数"数据，2025年金融行业大模型中标项目达587个，披露金额15.06亿元，较2024年的133个项目、2.4亿元实现 跨越式增长。 21世纪经济报道记者郭聪聪 2025年初，深度求索公司推出的AI大模型DeepSeek，凭借其开源、低成本与高性能的突出优势，迅速在金融业引发应用热潮，被视为继移动互联网之后， 对行业底层运作逻辑最具颠覆性的技术力量。 经过一年的发展探索，银行业的AI竞速，正在进入下半场。 21世纪经济报道记者从2025年度的招投标市场观察到，金融机构对大模型的采购与应用呈现爆发式增长。招投标数据显示，金融行业大模型中标项目数量同 比激增341%，披露金额飙升527%，百度、科大讯飞（002230）、火山引擎、中关村科金、阿里云等服务商纷纷进入了这场竞赛之中。其中银行业以近半数 的项目 ...

Core Viewpoint - The financial industry's digital transformation is accelerating, with data as the core driver, reshaping financial services and facing various challenges [2][4] Group 1: Digital Finance Concept - Digital finance is viewed as the "new infrastructure" of finance, supporting key areas like inclusive finance and green finance, representing a higher stage of financial technology [4][5] - The foundation of digital finance relies on data, which is essential for the digital transformation of various financial sectors [4][5] Group 2: Organizational Structure - The bank has established a dual committee structure to promote digital finance, with a Digital Finance Promotion Committee led by the president and a specialized AI Innovation Application Committee under the IT Committee [5][6] - The IT and big data departments are tasked with building a robust data platform to enhance resource efficiency and support frontline business needs [5][6] Group 3: Unique Positioning - The bank's mission is to leverage digital capabilities to practice inclusive finance and serve small and micro enterprises, maintaining a value system that prioritizes responsibility over profit [6][7] - The focus is on being "small and beautiful," emphasizing differentiation in customer base, technology, and product offerings rather than pursuing large-scale growth [6][7] Group 4: AI and Big Model Applications - The application of large models in finance is progressing slowly due to high compliance requirements, with current uses focusing on standardization and text-heavy tasks [7][8] - Examples include automating anti-money laundering processes, significantly improving efficiency and accuracy compared to manual methods [8][9] Group 5: Customer Interaction and Service - The evolution of customer interaction is moving towards a "one-sentence bank" model, allowing customers to express needs simply, with the system automatically identifying and fulfilling requests [9][10] - Innovations in customer service include intelligent customer support and service card systems that streamline user experience [9][10] Group 6: Data Management and Innovation - Emphasis is placed on foundational data work, which is crucial for enabling AI applications and fostering innovation [10][11] - The bank aims to solidify its data collection and management processes to support future technological advancements [10][11] Group 7: AI Governance - AI governance is complex, requiring a balance between regulation and innovation, with a focus on establishing safety boundaries and responsibilities [11][12] - The industry lacks mature governance experiences, suggesting the establishment of ethical committees and industry standards to ensure responsible AI development [12]