Yuanbao Inc. (YB) Q2 2026 Earnings Call Transcript

Ladies and gentlemen, good day and welcome to Yuan Bao Ying Second Quarter 2026 Earnings Conference Call. Today's conference is being recorded. At this time, I would like to turn the conference over to Ms. Stella Liu, Investor Relations and Strategy Associate Director. Please go ahead.

Thank you operator, please note that today's discussion will contain four looking statements made under the Safe Harbor provisions of the U.S. Private Securities Education Reform Act of 1995. Such statements are not guaranteed of future performance and are subject to certain risks and uncertainties, assumptions, and other factors. Some of these risks are beyond the complete control and could cause actual results to differ materially from those mentioned in today's press release and discussion. A general discussion of the risk factors that could affect Yuan Bao's business and financial results is included in certain filings of the company with the Securities and Exchange Commission. The company does not undertake any application to update this forward-looking information, except as required by law. During today's call, management will also discuss certain non-gov financial measures. For definition of non-gov financial measures and a reconciliation of GAAP to non-GAAP financial results. Please see the earnings release issued earlier today. Joining us today on the call from Yuanbao Senior Management are Mr. Wei Fong, our Chairman and Chief Executive Officer, and Mr. Wei Weng, our Chief Financial Officer. Mr. Fong will deliver his remarks in Chinese, followed by an English translation. All figures will be in R&B unless otherwise noted. We will conclude the call with a Q&A session. As a reminder, this conference is being recorded. In addition, a webcast replay of this conference call will be available on Yuan Bao's investor relations website. I will now turn the call over to our Chairman and CEO, Mr. Feng. Please go ahead.

Thank you. We will continue to invest AI technology, and continue to build a new product, and provide strong support for our industry.

Hello everyone, thank you for joining us today for our second quarter, 2026 Erdance Conference Call. In the second quarter of 2026, we maintained steady growth momentum, with both revenue and profitability improving in attendance. Total revenue for the quarter was 1.39 billion RMB, a year-over-year increase of 30.1%. Net income was 413.2 million RMB, up 35.6% year-over-year, with a net income margin of 29.7%. As of the end of the second quarter, our cash reserves totaled 5.16 billion RMB, providing solid support for our continued investment in AI technology, our product innovation in collaboration with ensuring carriers, and our ability to capture structural opportunities within the industry.

Thank you very much.

We continue to invest in strengthening our data and modeling capabilities. As of the end of the second quarter, our model matrix has exceeded 5,100 models, which are capable of analyzing more than 5,900 labels, covering the full range of scenarios from needs identification and product recommendations, to plan assistance, craving improvements in both service efficiency and the user experience in addition the 15 million US dollar share repurchase program we announced last quarter continues to progress steadily.

As of August 31, 2026 we had repurchased approximately 114,000 ADS for a total consideration of approximately 1.6 million US dollars reflecting our ongoing commitment to delivering capital returns to our Thank you very much. 鼓励增加覆蓋健康干预、特需医疗、创新药等的新型健康保险产品的供给。此前,国家金融监管总局在关于推动健康保险高质量发展的指导意见中明确指出,支持开发、液晒、带病体、罕见病等特定群体的三页医疗保险产品。

In recent years, commercial health insurance has been both developing opportunities a continuous stream of favorable policy signals. In March this year, the 2026 report on the work of the government explicitly called for accelerating the development of commercial health insurance and promoting the high-quality development of innovative drugs and medical devices in order to better meet the public's diverse needs for medical treatment and medication. In July, the State Council issued the National Health Plan for the 15th five-year plan period, encouraging an increased supply of innovative health insurance products covering areas such as health intervention, specialized medical care, and innovative drugs. Prior to that, the National Financial Regulatory Administration explicitly stated in its guiding opinions on promoting the high quality development of health insurance that is support the development of commercial medical insurance products targeting specific groups such as individuals with pre-existing conditions and the rare diseases against the spectrum we continue to deepen collaboration with insurer carriers and accelerate product iteration adhering to a strategy built on affordable pricing, comprehensive coverage, and quality service.

We are working to further improve the quality and expand the reach of the inclusive health protection by lowering enrollment barriers, enriching coverage benefits, and enhancing the service experience. 急诊,一般医疗到品质医疗的批度覆盖,让不同年龄、不同需求、不同预算的用户都能找到适合自己的保障方案。 今年6月,联保特医保系列医疗险迎来全面升级,等次升级围绕用户就医全流程的痛点,在好医、好药、好康库三个维度同步发力。

On the Prada side, Yuen Bao's super medical insurance series has improved over many years in response to consumer needs and now offers tiered coverage, many basic outpatient and emergency care, general medical treatment, and premium medical care so that users of different ages, needs, and budgets can all find a plan that fits in. In June this year, our super medical insurance series underwent a comprehensive upgrade focused on addressing pain points across users' full treatment journey. This upgrade advanced simultaneously across three dimensions, good doctors, good medicine, and good rehabilitation. It added access to special needs medical resources at public hospitals, expanded coverage for advanced drugs and medical devices, and introduced coverage for invitation rehabilitation for specific conditions, extending health protection further across the full cycle of treatment plus rehabilitation. Notably, even as we substantially upgraded our coverage, we stayed true to our principle of affordable pricing, improving quality without raising prices.

In addition to the COVID-19 pandemic, the COVID-19 pandemic has led to the COVID-19 pandemic.

This year, Yuan Bao has also been actively exploring new approaches to protection for individuals with pre-existing conditions, responding to national policies that encourages insurance for this population. In May, WEI partnered with insurers to launch complete Guardian Million RMB medical insurance, which with health disclosure requirements and covers general pre-existing conditions within the scope of benefits, effectively lowering the eligibility threshold for individuals with pre-existing conditions to obtain medical protections. On the critical unit side, WEI launched a no health disclosure version of Guardian Insurance million RMB critical illness insurance, further extending critical illness protection to more people who need it. From medical insurance to critical illness insurance, we continue to use product innovation to broaden access to health protection, directly addressing the real protection needs of the more than 400 million people in China living with pre-existing conditions.

This is an important part of the technology.

On the technology innovation front, we continue to advance the application of multi-modal AI technology across insurance business scenarios, building multi-modal understanding and analysis capabilities, tailored to the insurance industry. Insurance operations involve large volumes of unstructured information, including images, documents, videos, and tips.

By leveraging our multi-modal models and proprietary knowledge systems, Yuanbao enables AI to more accurately recognize, understand, and analyze this complex content, turning unstructured data that was once difficult to put to use into truly usable data assets. 形成结构化的理赔事件档案 并结合保审知识库 辅助责任分析和用户服务 以此来提升保释的理赔处理效率 审核此度一致性和用户的实验 目前相关能力已在医疗选理赔协助场景中落地 材料分类准确率达到95% 关键字段抽取平均准确率约94%

In clinical assistance service scenarios, we applied multimodal models to intelligently parse various types of materials, including medical records, diagnostic certificates, medical invoices, and examination reports, enabling document recognition, information extraction, fact consolidation, and review assistance. The system also across analyzes key information across materials from different sources to generate structured claims case files and combines with our insurance knowledge base to support liability analysis and user service, thereby improving insurance carriers with processing efficiency, consistency in review standards, and user experience. These capabilities have already been deployed in medical insurance clients' assistance scenarios, achieving a material classification accuracy of 95% and an average key field extraction accuracy of approximately 94%. Compared to the wind-endor solutions previously tested, our current solution shows improvement in both processing efficiency and accuracy. 。 8.

9. 10. 10. 10. 11. 11. 11. 12. We have a lot of information about AI in the world's world.

In user growth scenarios, we applied multi-modal models to intelligently understand and analyze business content. In this way, we effectively identify product information, expression characteristics, and user feedback across images, views, and text, and combine this with business performance data to support content analysis and a strategy optimization, improving user outreach efficiency in business operations. At the same time, we continue to build our insurance domain knowledge base. We structured core content covering product terms, medical knowledge, service workflows, clients rules, and historical cases into a professional knowledge foundation that underpins our AI applications. We also deeply integrated the knowledge base with multimodal models and agent technology, further enhancing the accuracy, interpretability, and business adaptability of AI in complex insurance scenarios and driving the evolution of insurance services from a traditional manual processing model toward an AI-assisted human-machine collaborative model. 共長大學五道口金融學院中國保險與養老金融研究中心發佈了2025年中國互聯網保險消費者重大報告。

商標室性供應對消費者的行為電源 深入泡汐了線上化 正能化及理性化的消費趨勢 並且全面梳理了行業演技與AI升級的方向 這不僅為行業高質量發展提供了重要的決策參考 更重要的是 我們將相關的調查 深度防護者 產品協同與創新服務升級中

Finally, turning to our industry influence and frontier practices, we continue to support industry development through professional research and drive insurance service upgrades through technology innovation. We also continue to strengthen our industry influence, growing on deep insights to capture structural opportunities in the industry. In May, together with the Research Center for China Insurance and Pension Finance at Tsinghua University's PBC School of Finance, we published the 2025 China Internet Insurance Consumer Insights Report. The report systematically tracked shifts in consumer behavior, took an in-depth look at the trends toward online, intelligent, and more rational consumption, and comprehensively We examined the direction of industry evolution and AI-driven upgrades. This provided an important reference for high-quality development across the industry.

And more importantly, we have channeled these insights deeply into our joint product innovation with insurers and service upgrades, further strengthening our overall competitiveness in internet insurance. 另一方面,我们将技术优势转化为卓越的运营效能,对线服务承诺,公司7月发布的2026年理赔半年报形式,智能资赔推动合作保释,实现了万年以下小额赔偿案件的结案失效,缩短了11%,同时,我们还推出了理赔追赞服务,增加了理赔调解服务, 确实做到了以科技提升服务温度

实现了用途体验和运营效率的双重提升 We are also translating our technological advantages into strong operating efficiency and delivering our service commitments According to our 2026 First Half Cleanse Report released in July Intelligent Cleanse Assistance helps partner insurance carriers shortened the clan settlement time for small clans under 10,000 RMB by 41%. At the same time, we introduced our clan's follow-up service and upgraded our clan's mediation service, bringing greater warmth to our service through technology and achieving a dual improvement in user experience and operating efficiency. 以及人工智能加速赋能千行百亿,保险行业正迎来科技创新,以保证需求升级交汇的新节度了。 如何让更多的消费者获得可负担有质量有温度的健康保证,即使行业发展的子弹硬体,也是元宝长期探索的方向。

面向未来,元宝将不断夯持AI技术底座,推动自能体、多模态、模型等技术,在保险全列落地,持续升级用户洞察、产品推荐、用户资讯,我们也将持续打磨、超医宝、全守护,切实进行,让保险满意图稳健经营、技术积累以及对用户需求的深刻理解。 Looking ahead, as the 15th five-year plan continues to advance, the multi-year healthcare security vision continues to improve.

The artificial intelligence rapidly empowers industry across the board. The insurance sector is entering a new phase where technology innovation and the rising protection needs converge. How to give more consumers access to house protection that is affordable, high quality, and compassionate is both a defining question for the industry and a direction Wenbao has long pursued. Going forward, we will continue to strengthen our AI technology foundation, bringing technologies such as AI agents and multimodal models into end-to-end use across our insurance business and upgrading core capabilities, including user insights, product recommendations, user consulting, and client assistance service. We will also continue to refine our inclusive product matrix, including super medical insurance and a complete guardian, Putting into practice our service philosophy of accessible, affordable insurance with high-class satisfaction, backed by stable operations, our accumulated technology capabilities, and a deep understanding of user needs, Yuan Bao will seize the opportunities presented by the insurance industry's high-quality development, drive innovation in commercial health insurance, support the development of the multi-tiered health care security system and create long-term value for consumers, partners, and shareholders. Now I'll turn the call over to our CFO, Ray Wan, to present our financial results for the second quarter of 2026. Thank you, everyone.

Thank you, Mr. Fa, and thank you everyone for joining today's earnings conference call. I'm pleased to share an overview of our second quarter's 2026 financial results. All figures are below are in R&B and let's note otherwise. We carry a strong momentum into the second quarter, delivering robust growth across our core business line while continuing to deepen AI integration across our operations. Total revenues for the second quarter reached $1.39 billion, representing a 30.1% year-over-year increase. This growth was primarily driven by strong increases in both our insurance distribution and system service revenues. Breaking down our revenue mix, revenues from insurance distribution services reached $457.4 million, a 30.4% year-over-year increase driven by a continued increase in the number of policies purchased through our platform, supported in part by our enhanced targeting marketing efforts. systems services revenue reached $881.9 million, up 22.8% year-over-year, reflecting continued improvements to our full consumer service cycle engine, which strengthened our ability to deliver more effective marketing, analytics, and customer-related services to our insurance carrier partners, along with expanded system services to both existing and new carrier We also generated $52.8 million revenue this quarter from advertising services, a new revenue stream we began offering this quarter. The newly provided advertising services, powered by the company's proprietary intelligent marketing platform, offered customers integrated intelligent marketing and traffic optimization solutions. Turning to expenses, total operating costs and expenses increased by 21.9% year-over-year to $941.3 million. Operations and support expenses were $97.8 million, up 139.1% year-over-year, primarily due to costs associated with our newly launched advertising services. Selling and marketing expenses increased 12.8% year-over-year to $679.3 million, consistent with our ongoing focus on consumer growth and retention. General administrative expenses were $64.8 million, up 36.3% year-over-year, primarily driven by higher professional service fees and increased personnel-related costs. Research and development expenses increased 21.7% year-over-year to $99.5 million, driven by continued headcount growth in our R&D organization and ongoing investment in technical capability building, which remain key pillars of our positioning as a technology-driven online insurance distributor. Below the operating line, income tax expense for the quarter was $59.9 million, compared with $11.1 million a year ago, primarily driven by a higher effective tax rate during the Net income for the quarter was $413.2 million, a 35.6% year-over-year increase, with a net income margin of 29.7%. Non-gap adjusted net income reached $431.2 million, up 32.6% year-over-year, with a non-gap adjusted net income margin of 31%. Our cash position grew during the quarter. As of June 30, 2026, our cash and cash equivalents, time deposits, restricted cash, and short-term investments totaled $5.16 billion, up 50.9% year-over-year and 8.8% from the end of first quarter. Net cash provided by operating activities in the quarter was $419.1 million. Before I close, let me provide an update on our shareholder return progress. Under our U.S. $15 million share repurchase program, as of August 31, 2026, we have repurchased approximately 114 ADS for a total consideration of approximately $1.6 million, in U.S. dollars, reflecting our continued commitment to returning capital to shareholders. To conclude, our second quarter results reflect the durability of our growth model and our ability to scale with discipline. Looking ahead, we will remain focused on high-quality growth, operational efficiency, and a healthy liquidity position, giving us the flexibility to invest in our strategic priorities while continuing to deliver long-term value for our shareholders. Thank you. And I would now like to open a call to Q&A. Operator, please go ahead.

Thank you, Management. We will now begin the question and answer session. To ask the questions on the phone, please press star one and one and wait for our name to be announced. For the benefit of all participants on today's call, if you wish to ask your questions to Management in Chinese, please immediately repeat your question in English. One moment for our first question. Our first question comes from the line of Amy Chen of Citi. Your line is open. Please go ahead.

Hi, this is Amy Chen from Citi. Thank you for the opportunity for asking this question. I have two questions. Our first is the name in Chinese and name in English. The first question is to ask the company how to look at the client's business through How do you think of licensed celebrities promoting insurance products for your live streaming? Would you adopt similar approaches going forward, and would this lead to more intense market competition in terms of online distribution of insurance products going forward? The second question is to ask the company, do you think the technology economy will be able to complete the entire contract? How will the company plan to try to catch the opportunity for the future? Do you think the Gentic economy would re-shape the insurance distribution industry? And how do you plan to capture the opportunities that are related to the Gentic payment? Thank you.

Thank you, Amy. I will answer your question first. 近期监管,新规明确的金融产品网络营销必须机构失牌,人员执政,执政民人直播带货本质上是这一合规框架下,利用公众影响力进行客户教育和流量转化的新形态。 We believe that the broadcast is a flow of flow, and the key is the key to the end of the market, which is the key to the end of the market. Therefore, we are trying to make the flow flow into the service of the product itself, and as opposed to the people of the company. At the end of the market, there will be a range of flow flow of light, and more of the quality of the product. Thank you. recent regulations have made clear that online marketing of financial products must be conducted through licensed institutions and by licensed individuals.

Licensed celebrity live streaming is essentially a new form of customer education and traffic conversion within this compliance framework. Leveraging public influencers which we believe that live streaming is merely a traffic entry point. The real barrier to insurance sales lies in back-end product interpretation and matching consumers with the right product. Therefore, we focus on following traffic into Yuan Bao's own service ecosystem rather than relying on external personal IPs. As for industry competition, celebrity participation may temporarily raise the intensity of traffic exposure and I intensified customer acquisition cost competitive competition for a standardized product however stringent regulatory restrictions have limited how far the economy monetization can go over the long term competition will continue to center our industry know-how proprietary data assets and ecosystem signature what matters more to us is that as this industry expense and the public awareness of insurance growth through live

streaming and other formats overall market penetration will continue to rise and we can capture shares through differentiated technology and the service. 第二个关于智能体的问题 我们认为AI agent可能会成为保险销售的新入口 就原宝而言 我们目前主要依托主流互联网平台进行分销 在可预计的未来 这些超级平台仍将是用户市场的绝对占据者 因此现有分销基本盘仍然稳固 当成用户使用AI以信息咨询为主 电商保险等交易转化链入仍然较薄弱,难以形成完整的投保闭环,但从中国互联网保险消费者洞察报告发现,AI智能体凭借企业辩解性与大模型技术的持续演进,显著降低了用户获取保险信息的门槛,在咨询体验上带来了实质性的效率提升。 对于标准化相对简单的保险咨询 AI的处理能力以可媲美人工 这一点无庸置疑 但我们也清醒地认识到 面对复杂深度的保障需求 与个性化方案设计 当前 AI仍可能存在幻觉 与准确性边界 技术成熟度尚在眼睛中 在落地层面 我们正积极将保险垂直

Do you believe AI agents could become a new entry point for insurance sales? Yuan Bao currently distributes primarily through mainstream internet platforms, and for the foreseeable future, these super platforms will remain the dominant holders of user time, so our existing distribution base remains solid. At present, user interactions with general purpose AI are confined to information inquiries. The conversion path for transactions such as e-commerce or insurance purchases remains relatively weak and is not yet sufficient to support a full insurance purchase journey. Even though our China Internet Insurance Consumer Insights Report found that AI agents, through their convenience and the continuous advancement for large language model technology have significantly lowered the barrier for users to access insurance information and many fully improved efficiency in the consultation experience. For standardized, relatively simple insurance inquires, there is no question that AI's capability now rival those of human agents. That said, we are also clear at that for complex in-depth coverage needs and personalized plan design, AI may still be subject to hallucinations and accuracy limitations, and the technology is still maturing. On the implementation front, we are actively integrating our insurance-focused vertical AI agent with external high traffic platform ecosystems, participating in the early stage development of this emergent traffic arena. We acknowledge that the application of AI agents in insurance scenarios is still at an early stage with both the ecosystem and our capabilities still maturing. At the same time, we firmly believe this represents an important shift in traffic entry points and how users interact. And it is an emerging trend and market that deserves close attention. Thank you. Next question.

One moment for our next question. The next questions will come from the line Ao Yue Xu from China Security. Your line is open, please go ahead.

短写的趋势有看到一些趋势性的变化,需求有一些趋势性的改善 然后第二个问题是海外拓展,因为去年下半年应该是拿了香港的保险经济牌照 就想问请关注update一下海外业务拓展的一个更新 Okay, my first question is in regards of the digital ad demand on major internet platform How is that impacting our customer acquisition and also, are you seeing any shifts in demand for short-term insurance? My second question in regards to oversee business progress. Could management team update on Hong Kong brokerage business since securing license last summer? which is the current level of ROI.

Now, the ROI is still in a relatively stable level, and it is not visible on the public demand demand. The short-term demand is still very large. The main goal is if the insurance products are related to the insurance products. The company will continue to apply insurance management, to increase the premiums, to match the different needs of the users, and to continue to expand the amount.

If customer scale were held constant, continuous optimization of our models and products would keep driving ROI higher. However, we continue to grow backspend and expand into new user segments, and external traffic competition dilutes ROI to some extent. Currently, our ROI is holding at a fairly stable level, and we have not seen any negative impact from software advertising demand yet. The demand base for short-term insurance products in China remains substantial. The key is whether we have coverage products that match that demand, we will continue to iterate on coverage terms and upgrade deductible rules to serve different tiers of user needs and keep unlocking incremental growth. We continue to explore various new business lives while strictly managing the ROI performance of new initiatives. Currently, we don't have any other new updates. Thank you. Next question, please.

The next question comes from Thomas Wang of Goldman Sachs. Your line is open. Please go ahead.

So firstly, on the advertising business, we want to see initial expense is quite high, but I want to see what's been revealed on median to long-term margin. And second point is on the trend in terms of our policies sold during the quarter and also the average ticket size. Thank you. In our advertising business, the entire margin is not a big deal, but it's still healthy. This is a new business to help our customers with better quality and quality. The second question is about our products. The products are not a big deal, but the overall趨勢 has no change in the past.

For the advertising new business right now, we don't disclose any details, but right now it's a healthy trend. As for the average price of our products, also we don't disclose any details in our quarterly report, but our momentum continues healthy. Thank you. Next question, please.

One moment for our next question Next up we have the line from Ling Tan from Hightong International Your line is open, please ask your question 感谢管理层给我这个提问的机会 我这边有两个问题想要请教一下 第一个问题的话是关于广告服务收入 可以请管理层具体讲一下这个广告服务项目具体指的是什么吗 第二个问题的话是关于AI 公司上半年持续落地了多Asian AI理赔体系 然后AI在理赔和保用户转化等等的环节都加大了投入

想请问一下目前AI相关的能力对公司的货客成本 保单质量 理赔成本的实际改善量化的效果如何 后续的话研发投入的资本开支有什么规划 I have two questions. Thank you, Management for taking my question. My first question is regarding the new advertising service revenue. What exactly is the new advertising service revenue this close in this quarter's revenue breakdown? And then my second question is regarding AI. In the first half, the company continued to roll out its multi-agent AI claim system and step up investment in AI across claim underwriting and user conventions. What quantified improvement has this AI capability actually delivered in customer acquisition costs, policy quality, and claim costs? And what's the CapEx plan for R&D investment going forward?

Thank you. for customers to provide a combination of digital marketing and high-quality and high-quality method. Currently, I think it's a little bit lower in this video.

Advertising services are a new service offering that was introduced during this quarter, specifically referring to services provided to insurance carriers. Our newly launched advertising services are built on our proprietary intelligent marketing platform, which provides clients with integrated intelligent marketing and traffic optimization solutions. Advertising services represented a small share of total revenue in this quarter. The first quarter, in our plan, the 15 scenarios we used, Multimodal models to intelligently parse a variety of materials, including medical records, diagnosis certificates, medical invoices, and examination reports, enabling document recognition, information extraction, fact consolidation, and review assistance. These capabilities have now been deployed in medical insurance claims assistance scenarios, achieving a documented classification accuracy of 95% and an average key field extraction accuracy of approximately 94%. As of the end of the quarter, our AI team continued to account for over 10% of the total workforce. Moving forward, we will continue to prioritize deep, consistent investments in our core technology capabilities. Thank you. Next question, please.

Our next question will come from the line of Tsing Tao Chen of CICC. Your line is open.

Please go ahead. 非常感谢管理层给我提问的机会 我有两个问题想请教 第一个问题就是公司在一季报里面宣布了派息和股份回购的计划 能莫请管理层介绍一下26年和27年两年派息和回购的政策的指引 第二个问题也是有关AI父母的问题 就是除了AI是着力提升新用户的转化效率之外 公司如何利用AI来提升老用户的流程率 So I have two questions. The first question is, in the first quarter, earning release, Yuen Bao announced a dividend and share buyback program. So could you provide some guidance on the dividend and buyback policies for 2026 and 2027? The second question is, besides focusing on improving new customer conversion efficiency, how does Yuen Bao leverage AI to enhance the the retention of existing customers. 根据实际回购情况评估是否需要增加或调整回购计划

Our future dividend payout ratio range will need to be assessed based on profitability and approved by the Board of Directions. While maintaining robust profit growth, we will remain committed to creating long-term value for our shareholders. Subject to our healthy cash flow and strong profitability, going forward, we plan to maintain and or grow shareholder dividends. And on the buyback side, we have announced a $50 million share repurchase program over a 12-month period. Moving ahead, we will evaluate whether to extend or adjust the program based on how it is actually executed. 所需的曝光次数更重要的是在曝光后通过对用户行为的深度理解和运营 那另一方面如本季度提到公司利用多模态技术协助保私提升理赔处理效率和用户体验 公司也持续建设保险领域专业知识库 形成了支撑AI应用的专业知识基础 On one hand, we use our AI engine to optimize user conversion and retention directly, both before user exposure, for example, by reducing the number of impressions needed to convert a user, and more importantly, after exposure. On the other hand, as mentioned, we use multimodal technology to help insurers enhance claims processing efficiency and the overall user experience. We also continue to build out our proprietary insurance knowledge base, forming the professional knowledge foundation that supports our AI applications. By converting our technology advantages into superior operational efficiency, we can better deliver our service commitment and enhance the consumer's experience, which in turn improves user retention and drives repeat repurchase. Next question, please.

One moment for our next question. Our next questions will come from the line of Jackie Sun from Shen Wan Hong Yuan. Please go ahead.

Thanks for the opportunity. This is Jackie Sun from Shen Wang宏園 Securities. I have two questions. With the administrative measures of the online marketing of financial products scheduled to take effect in September, What does management see as the greatest challenge in the company's marketing compliance transition? How do you plan to ensure a smooth transition? The second question is, with the company's current ample cash reserves, how does the management assess the sustainability of its future dividend policy? Is management considering establishing a regular dividend mechanism tied to net income or free cash flow?

Thank you, Jackie. 这两个问题我来回答吧 第一个是关于金融产品网络营销管理办法 那截至目前该管理办法的出台 为对我们的后客模式产生任何实质性的影响 那这主要得益于两方面 一是我们一贯坚持合规先行 以搭建起完善的营销行为管理体系 并配套实施强制性合规审查流程 二是我们的核心业务逻辑 适于持牌保险公司的深度绑定 To date, the introduction of this regulation has not had a material impact on our customer acquisition model.

This is mainly due to two factors. First, we have always prioritized compliance and have already built a comprehensive marketing management system paired with a mandatory compliance review process. Second, our core business model is deeply tied to licensed insurance carriers, and all of our product terms go through strict regulatory approval for filing procedures to ensure our operations remain compliant. At the same time, we maintain ongoing proactive communication with regulators, striving to keep our corporate strategy closely aligned with the latest policy direction. 关于分红政策刚才也提到,在保持利润稳健增长的情况下,我们将继续助力于为股东创造长期价值,在未来我们计划维持或提高股东分红政策。 While maintaining robust profit growth, we will remain committed to creating long-term value for our shareholders. Subject to our healthy cash flow and strong profitability going forward, we plan to maintain or grow shareholder dividends. Thank you. Next question.

Please hold for our last question. Our last question comes from the line of Liao Yuan of CITIX.

Please go ahead. 哎, 感谢光丽森接受我们提问两个问题, 第一个是关于我们的AI, 那我们之前也上线了基于多Agent 协同的消费者智能保险咨询规划的一个工具, 那我想问当前这个工具的一个这个渗透的一个情况以及对我们效率带来多少的一个改善, 那在这个理赔辅助还有客户质检, 这些弊端的一个场景的话, AI Agent 带来的一个效率的提升,有没有数据可以分享。 那第二呢,就是我们看到公司的这个现金储备也是非常充足的, 那同时也是宣告了股东回报计划。 那后续的这个资金的使用计划, 我自己翻译一下。 Thanks management for taking my questions. I have two questions. The first one is about your AI. so could you share the penetration and attach rate of your multi-agent AI insurance advisor too and the efficiency if in gain from AI agent in B2B client support and quality monitoring scenarios and second question is about your creation giving the strong cash position and last quarter shareholder Shareholder Return Program. So could management share more detail about your future capital allocation plan?

Thank you. 从长期价值来看AI辅助或替代完全替代人工科服 理赔协助可以优化用户的服务体验 拉高用户LTV自动化建模和开发流程也能很快的加快模型迭代的速度 7月我们发布了2026理赔半年报 显示依托自研的AI辅助理赔系统 通过事件整理和预审及结论解读三大智能体的高效系统配合智能助赔功能复能合作嫌弃,我们协助合作保私实现了万元以下小额理赔案件结案时效缩了41%了,同时通过推出理赔追办,持续升级理赔调节等增值服务, 确实做到了以科技提升服务温度

实现了用户体验与运营效率的双重提升 By combining our knowledge base with multi-modal models and agent technology We have further improved AI's accuracy, interpretability and business adaptability In complex insurance scenarios Driving the evolution of insurance services from traditional manual processing toward an AI-assisted human-machine collaborative model. From a long-term value perspective, using AI to assist or replace manual customer service and client support can improve the user service experience and raise user LTV. Automating our modeling and development workflows also accelerates the pace of model iteration. According to our 2026 first half clients report we released in July, this was driven by our proprietary AI-assisted clients assistant, which relies on three AI agents working together efficiently for case organization, preliminary liability review, and conclusion interpretation. It works alongside our intelligent clients assistant feature to empower our insurance carrier partners, helping insurance carriers shorten client settlement time for small clients under 10,000 RMB by 41%. At the same time, by launching value-added services such as client follow-ups and continuously upgrading others such as client mediation, we brought more warmth to our service through technology, achieving improvements in both user experience and operating efficiency. 关于资金,我们打算继续保持严谨的资本配置策略,在现有的业务投资,潜在的新增长机会和选顺的流动性管理以及股东回报之间做好平衡。 We intend to maintain a disciplined capital allocation strategy, striking a balance among investments in our existing business, potential new growth opportunities, prudent liquidity management and shareholder return. Thank you.

And that concludes the question and answer session. I would now like to turn the conference back over to management for any additional or closing comments.

Thank you once again for joining us today. If you have any further questions, please feel free to contact us directly for presented financial communications. Our contact information for IR in both China and the US can be found in today's press release. Have a great day.

Today's conference call, thank you for your participation. You may now disconnect your lines.