车险AI大模型

Search documents
中国平安披露最新“五智”战略:AI IN ALL
2 1 Shi Ji Jing Ji Bao Dao· 2025-08-28 08:09
Core Viewpoint - Ping An's mid-term performance report highlights its strong foundation and leading advantages in financial technology, driven by its "AI IN ALL" strategy aimed at integrating AI across the entire value chain of finance, healthcare, and elderly care [1][2]. Group 1: AI Strategy and Development - Ping An's "Five Intelligence" strategy includes intelligent marketing, service, operations, management, and business, emphasizing the comprehensive application of AI [1]. - The company has established a full system of AI capabilities, including data, algorithms, scenarios, and computing power, supported by massive data accumulation [1][2]. - Ping An's database contains 30 trillion bytes of information, covering nearly 247 million individual customers, positioning it among the world's leading data accumulators [6][7]. Group 2: AI Technology and Applications - The accuracy of Ping An's algorithms is industry-leading, with self-developed technologies in key areas such as vision, hearing, and big data, resulting in a total of 460 published papers [2][12]. - The company has developed 67 proprietary domain-specific models, with 14 new models added in the first half of the year, enhancing its capabilities in underwriting, pricing, claims, and operations [2][12]. - Ping An has achieved zero-code development for all employees, enabling rapid deployment of intelligent applications, with 23,000 new intelligent applications launched in the first half of the year, covering 22% of its workforce [2][15]. Group 3: Performance Metrics and Impact - As of June, Ping An's large model has been called upon 818 million times, with over 650 applications across various scenarios [3][15]. - In the life insurance sector, the "AI Trainer" covers 31% of new individual insurance agents, effectively boosting activity rates and productivity [3][17]. - In healthcare, the AI product system launched by Ping An Good Doctor achieves over 95% accuracy in medical document assistance and over 90% improvement in chronic disease management [3][19]. - The car insurance AI model enhances pricing accuracy and optimizes business quality, contributing to a 0.3 percentage point reduction in claims costs over the past three years [3][21].