Core Insights - Baidu has launched the world's first commercially available "self-evolving" superintelligent agent, named Baidu Famo, at the 2025 World Conference [1] - China CITIC Bank has introduced the "AlphaMo" intelligent risk control project, which incorporates Baidu Famo, marking its first application in the financial sector [1] Group 1 - Baidu Famo represents a significant advancement in artificial intelligence technology, showcasing the company's commitment to innovation in the field [1] - The collaboration with China CITIC Bank highlights the potential for AI applications in financial services, indicating a growing trend of integrating advanced technology into traditional banking [1]

中信百信银行以数字风控为基础，携手百度启动"AphaMo"智能风控项目。该项目应用了伐谋、千 帆、文心大模型等百度最前沿的AI技术，建设"挖掘智能体""模型智能体""策略智能体"三类风控智能 体，覆盖风控架构中挖掘、模型、策略三个层次。同时，还将发挥AI擅长处理复杂数理问题、持续自 我进化的特点，打造挖掘、模型、策略的"AI员工"，既能解决风控人员短缺问题，又能解决人的经验局 限性问题，有望为银行业沉淀出可复制的智能风控创新解决方案。 据了解，当前"AphaMo"智能风控项目中的"挖掘智能体"已取得超预期进展。"挖掘智能体"由中信 百信银行专业风控专家进行训练指导，利用百度伐谋擅长处理复杂数理问题、持续自我进化的特点， 7*24小时不间断地在海量数据中进行风险特征挖掘，持续学习、迭代、进化，在短时间内达到了专业数 据挖掘工程师的水平，特征挖掘效率提升100%，风险区分度也提升了2.41%。 百度创始人李彦宏介绍，伐谋的理念借鉴自进化算法，相当于模拟生物界几亿年的进化过程并压缩 至几天甚至几小时，从而发现"过去人类从未发现过的全局最优解"。 11月13日举办的百度2025年世界大会上，百度正式发布全球首个可商用 ...

Core Insights - Baidu has launched the world's first commercially available "self-evolving" super intelligent agent, named Baidu Famo, at the 2025 World Conference, marking a significant advancement in AI technology for financial innovation and inclusive finance [1] - The "AphaMo" intelligent risk control project, implemented by Citic Baixin Bank, is the first application of Baidu Famo in the financial sector, showcasing the integration of AI in enhancing digital financial services [1] Group 1: Project Overview - The "AphaMo" project utilizes Baidu's cutting-edge AI technologies, including Famo, Qianfan, and Wenxin large models, to create three types of risk control intelligent agents: digging, modeling, and strategy agents [1] - The project aims to address challenges in the banking industry, such as insufficient feature extraction and a shortage of specialized talent, by leveraging AI to automate and enhance risk control processes [1][3] Group 2: Performance Metrics - The "digging agent" has achieved significant progress, with a 100% increase in feature extraction efficiency and a 2.41% improvement in risk differentiation, reaching the level of professional data mining engineers [2] - The "modeling agent" has demonstrated an 83% improvement in efficiency, completing tasks in 0.5 days compared to 3 days for junior model engineers, and its models have surpassed the performance of entry-level personnel by 7% in KS value [3] Group 3: Strategic Implications - The implementation of intelligent agents in the "AphaMo" project effectively addresses the talent bottleneck in risk control, allowing for automated and efficient feature extraction and model development [3] - The project enhances risk identification capabilities, enabling the bank to serve a broader customer base while maintaining controlled risk levels, thus supporting the expansion of inclusive financial services [3] - The collaboration between Citic Baixin Bank and Baidu represents a new paradigm in the integration of AI technology with banking risk control, aiming to improve the efficiency and quality of digital inclusive finance [3]