Core Insights - The conference highlighted that AI technology is profoundly transforming the financial industry, accelerating the pace of application and becoming a key driver for restructuring business models, enhancing operational efficiency, and optimizing service experiences [1]. Group 1: Impact of AI on Financial Institutions - AI technology is impacting financial institutions across all fronts: front-end customer interactions through intelligent customer service systems, mid-end product development and service integration, and back-end internal management leading to flatter and more agile organizational structures [2]. - The performance of domestic large models has significantly improved due to breakthroughs in computing power, data, and algorithms, with a shift in computing demand from training to inference, indicating a trend towards low precision, large memory, and high bandwidth in AI chip development [2]. Group 2: Differentiated Development Paths - Large financial institutions should focus on business restructuring, process reengineering, and organizational transformation driven by technology, leading to new products, models, and business formats [2]. - Small and medium-sized financial institutions are encouraged to pursue differentiated and specialized digital transformation paths based on their resource endowments to improve the technology investment-output ratio [2]. Group 3: Collaborative Ecosystem Development - Financial institutions are urged to strengthen open collaboration in ecosystem construction, with large institutions taking the lead in exploring multi-scenario applications of AI technology, while smaller institutions should adopt an open cooperative attitude around high-frequency business scenarios [3]. - The China Academy of Information and Communications Technology (CAICT) aims to facilitate the deep application of large models in the financial sector by collaborating with various industry stakeholders [3]. Group 4: Risk Management and Evaluation - AI technology vulnerabilities may introduce unpredictable new operational risks for financial institutions, necessitating the inclusion of AI risk in comprehensive risk management frameworks [3]. - An evaluation system covering the entire lifecycle of large models is essential to help institutions clarify optimization directions and allocate resources effectively, as traditional annual assessments are insufficient to keep pace with technological iterations [3].

● 本报记者 刘英杰 欧阳剑环 9月15日，由华鑫证券主办的"先机——金融业'人工智能+'跃升计划 暨华鑫证券智能投研平台'鑫i'新品 发布会"在上海召开。 与会专家表示，人工智能（AI）技术正深刻改变金融行业生态，未来应用步伐将继续提速，成为推动 金融业务模式重构、运营效率提升和服务体验优化的重要引擎。 AI技术正全面影响金融机构 国家金融与发展实验室副主任、上海金融与发展实验室主任曾刚在会上表示，AI技术的不断突破正在 从前、中、后端全面影响金融机构。前端"面客"场景可通过智能客服系统实现自然语言处理能力的飞 跃；中端产品开发和服务集成环节可提升效能；后端内部管理场景则推动组织架构向扁平化、敏捷型转 变。 中国信息通信研究院副总工程师何伟在演讲中表示，当前大模型技术在算力、数据、算法等核心要素方 面持续突破，国内大模型性能显著提升。算力需求正从训练侧向推理侧转移，低精度、大显存、高带宽 成为AI芯片的发展趋势。垂域大模型因更贴合行业需求而不断涌现，正成为推动金融业务发展的重要 力量。 谈及应用落地，两位专家都强调了差异化发展路径的重要性。 合力推动AI技术场景落地应用 与会专家一致认为，金融机构要在生态建 ...