近日，2026巴伦中国价值榜单揭晓，元保集团入选“中国行业先锋30强”，该榜单重点关注企业在细分领域的技术应用与商业模式创新。
元保此次入选，为观察保险中介的技术转型提供了一个案例。随着线上投保逐渐普及，互联网保险平台需要解决的问题更加具体，如何帮助消费者选择适合的产品，如何降低保险公司的获客和服务成本，如何让投保时承诺的保障在理赔环节得到兑现等问题都需要得到回应。
这些能力也构成了理解元保业务的主要线索。作为连接消费者与保险公司的互联网保险中介，元保以技术优势驱动保险服务升级，其经营表现与产品销售、技术服务和用户运营密切相关。
AI如何进入保险日常业务
元保的业务覆盖产品推荐、线上投保、保单管理、理赔协助及售后服务，平台销售的保险产品由合作保险公司承保。对消费者而言，元保提供购险与服务入口。对保险公司而言，它承担了连接用户、辅助销售和持续服务的部分职能。
这一分工体现在公司的收入结构中。财报显示，2026年上半年，元保实现收入27.08亿元，同比增长32.7%，净利润延续稳健表现。而元保的稳健增长来自产品服务升级和技术深度应用。
中信建投证券曾在研报中将技术引擎视为其差异化竞争力，认为数据积累与模型迭代能够提高获客、转化和服务的自动化水平。
据元保披露，截至2026年二季度末，公司消费者全周期服务引擎已拥有5100余个模型，可分析5900余个标签，应用于需求识别、产品推荐及理赔协助等场景。这套系统将用户、产品与营销相关信息结合，用于支持不同环节的业务判断。
客户咨询是技术进入日常业务的一个具体场景。保险咨询往往涉及保障范围、健康告知、除外责任和理赔条件，同一个问题在不同产品下可能对应不同答案，需要结合具体条款作出解释。
针对这类需求，元保将大语言模型与保险专业知识库结合，应用于产品咨询、保单查询和理赔协助。据公司披露，其知识库收录了数千款保险产品及条款信息，并将核保规则、服务流程等内容结构化。智能客服通过检索增强生成技术，结合知识库中的产品规则组织回答。目前，该系统已覆盖售前咨询、保单查询、售后理赔等十余类场景，提供全天候服务。
另一项应用是处理理赔材料。元保在财报中披露，公司利用多模态模型识别病历、诊断证明、医疗票据和检查报告，提取并交叉核对信息，形成结构化档案。在医疗险理赔协助场景中，公司披露的材料分类准确率为95%，关键字段提取平均准确率约为94%。这些功能承担了部分原本需要人工逐项阅读、整理和核对的工作。
产品怎么选，理赔怎么办
技术应用的另一端，是消费者对保险产品与服务的需求变化。
清华大学五道口金融学院中国保险与养老金融研究中心与元保近日共同完成的《中国互联网保险消费者五年回顾与洞察》，汇总了2021年至2025年累计约1.67万人次的问卷调研。报告显示，在2024年的受访消费者中，63%存在产品选择困难，这一问题首次超过理赔担忧，成为购险时的主要困扰。
产品供给增加，让消费者拥有更多选择，也增加了比较难度。对互联网保险中介而言，识别这些需求并反馈给保险公司，是参与产品设计的重要基础。
元保近期的产品调整，主要围绕医疗保障范围和投保门槛展开。按照公司披露，其与保险公司合作推动医疗险升级，覆盖免赔额、先进医药保障以及就医、康复服务等内容。同时，公司推出面向带病人群的产品，将此前较难获得充足保障的人群纳入保障范围。
产品能否转化为实际保障，还需要在理赔环节得到检验。对于用户，提交什么材料、是否需要补充证明、如何理解审核结果，都会影响一次理赔的体验。
元保披露的理赔系统，将相关工作分配给不同的智能体。事件整理智能体识别票据、诊断证明等材料，核责智能体辅助责任审核与规则匹配，结论解读智能体则将条款逻辑和审核依据转化为便于用户理解的说明。
公司公布的《2026理赔半年报》提供了具体案例。一位用户因头痛产生门诊支出，在线提交申请后，3.3分钟完成审核赔付，获赔300元。另一位用户申请住院医疗理赔，6.2分钟完成结案，获赔3575.2元。公司还披露，上线智能助赔后，截至2026年6月，万元以下小额理赔案件结案时间缩短41%。
对于元保这样的互联网保险平台，技术投入的长期价值，最终要体现为更清晰的产品匹配、更有效的服务响应，以及更稳定的用户关系。随着行业竞争深入，获客效率与投保后的服务质量，将共同决定平台的增长能走多远。 （作者｜蔡鹏程，编辑｜刘洋雪）
更多对全球市场、跨国公司和中国经济的深度分析与独家洞察，欢迎访问 Barron's巴伦中文网官方网站
本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
In recent days, the 2026 Baron China Values List, which focuses on technological applications and business model innovations in the subdivisions of enterprises, has come to light with the selection of the China Industry Pioneer 30.
This election provided a case study of the technological transformation of insurance intermediaries. As online coverage becomes more widespread, the issues that Internet insurance platforms need to address are more specific, such as how to help consumers choose suitable products, how to reduce the cost of acquiring customers and services from insurance companies, and how to allow insurance guarantees promised to be met at the settlement stage.
These capabilities also constitute the main clue to understanding the guarantee business. As an Internet insurance broker linking consumers to insurance companies, MES drives the upgrading of insurance services with technological advantages, whose performance is closely linked to product sales, technical services and user operations.
How does AI get into insurance routines?
It covers product referrals, online insurance, policy management, settlement assistance and after-sales services, and insurance products sold on the platform are covered by cooperative insurance companies. For consumers, they provide access to purchase insurance and services. For its part, the insurance company performs part of its functions of connecting users, supporting sales and continuing services.
This division of labour is reflected in the corporate income structure. According to the financial statements, in the first half of 2026, the guarantee achieved income of 2,708 million yuan, an increase of 32.7 per cent over the same period, with a steady continuation of net profits. The robust growth of the guarantee comes from product service upgrades and technology depth applications.
In its development, CICAS has considered the technology engine as its differential competitiveness and that data accumulation and modelling over time can increase the level of automation of customers, transformations and services.
By the end of the second quarter of 2026, the MEA disclosed that the company had more than 5,100 models for a full-cycle consumer service engine that could analyse more than 5,900 labels for use in such scenarios as needs identification, product referral and settlement assistance. The system combines user, product and marketing-related information to support different segments of business judgement.
Client counselling is a specific scenario for technology entering daily operations. Insurance counselling often involves coverage, health information, exclusionary liability and compensation conditions, and the same question may vary under different products and needs to be explained in relation to specific provisions.
In response to such needs, the MES combines the large-language model with the insurance expertise base for product counselling, policy queries and claims settlement assistance. According to company disclosures, its knowledge base captures thousands of insurance products and information on terms and conditions and structures the content of safeguards rules, service processes, etc. Smart customer service enhances generation technology through retrieval, combining product rules in the knowledge base. Currently, the system has covered more than 10 types of scenes such as pre-sale counselling, policy queries and after-sales settlements, and provides round-the-clock services.
Another application is the processing of claims. The guarantee is disclosed in the financial statements, and companies use multi-mode models to identify medical records, diagnostic certificates, medical papers and inspection reports, extract and cross-check information to form structured files. In the medical insurance assistance scenario, the company disclosed an accurate classification rate of 95 per cent for material and an average accuracy rate of approximately 94 per cent for key field extractions. These functions carry out some of the tasks that would have required manual reading, collation and verification.
What about the product? What about the settlement?
The other end of technology application is changing consumer demand for insurance products and services.
The China Centre for the Study of Insurance and Pension Finance of the University of Qinghua ' s five-year Review and Insight of China ' s Internet Insurance Consumers, which was jointly completed by the China Centre for the Study of Insurance and Pension Finance of the University of Qinghua and China in conjunction with China ' s recent days, compiled a cumulative survey of approximately 16,700 people between 2021 and 2025. The report showed that 63 per cent of the consumers interviewed in 2024 had difficulties in product selection, a problem that, for the first time, exceeded the concerns of compensation and was a major problem in the acquisition of insurance.
Increased supply of products, giving consumers more choices, also makes it more difficult. For Internet insurance intermediaries, identifying these needs and providing feedback to insurance companies is an important basis for participation in product design.
The most recent product adjustment of the MES is mainly centred on the coverage of health care and the insurance threshold. As disclosed by the company, it works with insurance companies to promote the upgrading of medical insurance coverage, covering exemption amounts, advanced medical security and access to medical care and rehabilitation services. At the same time, companies have introduced products for people living with disease, which have included those who had previously had less access to adequate security.
The transformation of products into physical security also needs to be tested at the compensation chain. For users, what materials are submitted, whether additional proof is required, and how the results of the review are understood will affect the experience of a settlement.
The claims settlement system disclosed by MES assigned the related work to different AI agents. The events were sorted out in the form of AI agent identification papers, diagnostic certificates, the verification of AI agent subsidiary liability matched the rules, and the interpretation of AI agent translated the logic of the clause and the basis of the audit into a user-friendly statement.
Specific cases were provided in the company ' s biannual 2026 settlement newspaper. One user incurred out-patient expenses due to headaches and received $300 after 3.3 minutes of review and payment after online submission of the application. Another user applied for medical settlement for hospitalization, which was completed in 6.2 minutes, and was awarded $3575.2. The company also disclosed that as of June 2026, the completion time of small settlement cases of less than $10,000 had been reduced by 41 per cent.
The long-term value of technology inputs for an Internet-based insurance platform such as meso will ultimately be reflected in clearer product matching, more effective service response and more stable user relationships. As industry competes, the efficiency of recipients and the quality of services after insurance will collectively determine how far the platform can grow. YOON-YO'S EDITION
More in-depth analysis and exclusive insight into global markets, transnational corporations and the Chinese economy, welcome to Barron's Chinese-language website
Auto-fetched and translated by AI Atlas. All rights belong to the original author.