执业证书编号：S0740515080001 Email：kangyw@zts.com.cn AI 应用重塑流量生态，推动营销需求趋势向上 ——2026 AI 应用系列深度报告（一） | 基本状况 | | | --- | --- | | 上市公司数 | 130 | | 行业总市值(亿元) | 19,217.14 | | 行业流通市值(亿元) | 17,788.46 | 动，传媒攻守兼备》2025-12-16 透视 AI 算力需求》2025-11-27 传媒 证券研究报告/行业深度报告 2026 年 02 月 26 日 | 评级： | 增持（维持） | | 重点公司基本状况 | | | | | | | | | | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | | | | 简称 | 股价 | | | EPS | | | | | PE | | | 评级 | | 分析师：康雅雯 | | | | （港元/元） | 2023A | 2024A | 2025E ...

Core Insights - Douyin is launching a long-form content feature by the end of 2025, allowing creators to publish in-depth articles, marking a shift from short video content to longer, more substantial narratives [1][2] - The introduction of AI technology is central to Douyin's strategy, enabling efficient content generation and distribution, particularly in the realm of news and long-form articles [1][6] Group 1: Transition from Short to Long Content - Douyin's long-form content feature supports articles ranging from 300 to 8000 words, aiming to attract quality creators and enhance user engagement through deeper narratives [2][3] - The competition for user attention is intensifying, with platforms like Xiaohongshu also exploring long-form content, indicating a broader industry trend towards deeper engagement [2][3] - Douyin's move reflects a recognition that solely relying on short videos may not sustain future growth, prompting a search for new content forms [3][5] Group 2: AI Integration in Content Creation - Douyin is leveraging AI to enhance content production, with features that allow for real-time news aggregation and summarization, thus improving user experience [6][7] - The AI-driven news content will compete with original long-form articles within Douyin's ecosystem, raising questions about the quality and reliability of AI-generated information [7][8] - The rise of AI-generated content poses challenges for traditional media, as it blurs the lines of accountability and trust in news dissemination [8]

2月26日，微信员工@客村小蒋 发文称，最近微信还推出新功能：长按聊天中的任一图片、视频或文 件，可以看到"查看该图片/视频/文件包含的聊天"的选项，点击后就能看到它在多少个聊天中被使用的 情况。 可以怎么尝试解决呢？ 1. 点开微信设置里的【存储空间】页面，加载完成后，观察总占用，以及各个部分的占用； 据介绍，界面上还会出现说明："同一图片、视频和文件在多个聊天使用时，不会重复占用存储空间。" 据微信员工@客村小蒋，几个月前和同事在关于微信占存储空间的播客里聊过： 当用户去转一张图片、转一个视频的时候，转完了以后长按这个图片视频，会在他分享菜单里面新增一 个入口，让用户能够查看到，当前这个内容被多少个聊天使用过，他就可以看到之前可能是被5个聊天 使用，我转发多3个聊天，这个5可能就会变成8。 我们是尝试从UI表现上让用户知道，原来我转发多次，文件没有变多，只是被用到的次数变多了，有 点类似于这个文件实体被不同的聊天引用，引用的次数变多了，不代表我存储这个文件的体积会变大。 这个也让用户更放心，它从UI上可以看到，我的文件没有变多，只是被引用的次数变多了而已。 据其介绍，用户还可以用这个功能来查看，图片、视频 ...

Core Viewpoint - The article discusses the phenomenon of users receiving friend requests on WeChat from strangers using virtual WeChat IDs generated by AI platforms, raising concerns about privacy and data security [1][2]. Group 1: AI Platforms and User Experiences - Users have reported receiving friend requests from strangers via AI-generated WeChat IDs, leading to feelings of being "betrayed" when they discover the source of the contact [2]. - The AI platform "豆包" provided a user with multiple virtual WeChat IDs, one of which matched a real user's ID, while the platform emphasized the importance of privacy and stated that real IDs cannot be generated [6][7]. - Other AI platforms, such as Deepseek and 千问, also allow users to request the generation of virtual WeChat IDs, with varying degrees of success in matching real users [6][7]. Group 2: Privacy Concerns and Expert Opinions - Experts suggest that the ability of AI to generate IDs that match real users could stem from publicly available information or the AI's training data, which may include existing WeChat IDs [10]. - 聂再清, a researcher from Tsinghua University, indicated that the AI models might generate plausible WeChat IDs based on statistical probabilities, but if a platform can consistently generate a high percentage of real IDs, it warrants further investigation into their methods [10]. - The WeChat customer service advised against using personal information in WeChat IDs, highlighting the importance of privacy in the context of AI-generated content [11].

微信员工@客村小蒋 发文：最近微信还有个新的小能力，长按聊天中的任一图片、视频或文件，可以 看到「查看该图片/视频/文件包含的聊天」的选项，点击后就能看到它在多少个聊天中被使用的情况。 界面上还有个说明： 2月26日，微信员工@客村小蒋 发文称微信能看图片使用次数了，随即该话题登上微博热搜榜一。 "同一图片、视频和文件在多个聊天使用时，不会重复占用存储空间。" 几个月前，在和产品、开发同学在关于微信占存储空间的播客里，聊到过这个事情： "近期我们也在规划，当用户去转一张图片、转一个视频的时候，那当他转完了以后长按这个图片视 频，我们可能会在他分享菜单里面新增一个入口，让用户能够查看到，当前这个内容被多少个聊天使用 过，他就可以看到之前可能是被5个聊天使用，我转发多3个聊天，这个5可能就会变成8。 我们是尝试从UI表现上让用户知道，原来我转发多次，文件没有变多，只是被用到的次数变多了，有 点类似于这个文件实体被不同的聊天引用，引用的次数变多了，不代表我存储这个文件的体积会变大。 这个也让用户更放心，它从UI上可以看到，我的文件没有变多，只是被引用的次数变多了而已。" 我还观察到，有些朋友会用这个功能来查看，图片 ...

2月26日中午，#微信能看图片使用次数了#话题登顶微博热搜榜，引发网友热议。 2月26日，微信员工@客村小蒋发文称，最近微信还有个新的小能力，长按聊天中的任一图片、视频或 文件，可以看到"查看该图片/视频/文件包含的聊天"的选项，点击后就能看到它在多少个聊天中被使用 的情况。 他还透露，同一图片、视频和文件在多个聊天使用时，不会重复占用存储空间。这个功也能查看，图 片、视频或者文件都转发过给谁。 2月26日，微信员工@客村小蒋发文称，最近微信还有个新的小能力，长按聊天中的任一图片、视频或 文件，可以看到"查看该图片/视频/文件包含的聊天"的选项，点击后就能看到它在多少个聊天中被使用 的情况。 他还透露，同一图片、视频和文件在多个聊天使用时，不会重复占用存储空间。这个功也能查看，图 片、视频或者文件都转发过给谁。 不过对此功能，有网友质疑"把隐私做新功能"，还有网友吐槽"无聊的功能增加了"。 此外，微信最近还更新"面对面收照片和文件"的功能。微信官方表示，手机不同，没加好友，不用流 量，也能快速收发原图。 责任编辑：王翔 2月26日中午，#微信能看图片使用次数了#话题登顶微博热搜榜，引发网友热议。 不过对此功能 ...

（全球TMT2026年2月26日讯）今日要点：字节跳动估值已达5500亿美元；SK海力士将在韩国投资150亿美元新建半导体工厂； 三星电子发布Galaxy S26系列；英伟达业绩持续大幅增长年营收突破2000亿美元；日本反垄断机构对微软展开调查；中国手机行 业预计迎来全面涨价潮。 字节跳动估值已达5500亿美元 三星电子(Samsung Electronics)周三发布最新旗舰智能手机Galaxy S26系列，其中两款机型较前代涨价100美元，当前行业正努力 摆脱全球存储芯片短缺的影响。在AI基础设施快速扩张引发全行业存储芯片短缺的背景下，三星主推新机型升级的AI功能与隐 私显示屏。 投资机构General Atlantic正计划出售其持有的字节跳动部分股权，这笔潜在交易对这家中国社交媒体巨头的估值高达5500亿美 元。此次股权剥离是自特朗普政府于今年1月批准TikTok美国业务出售以来的首笔重大交易。这一最新估值较去年员工股票回购 时逾3300亿美元的定价实现了66%的显著增长。 抖音全年清退违规达人超27万名 抖音电商安全与信任中心发布的首份年度报告显示，过去一年，抖音电商着重治理虚假宣传，全年累计处置违 ...

前言 2025年，我国国内生产总值(GDP)首次突破140万亿元，达到1401879亿元，比上年增长5.0%。 消费需求平稳增长。消费品以旧换新等提振消费政策协同发力，消费需求持续释放，消费市场保持平稳增长。 2025年，最终消费支出对经济增长贡献率为52.0%。 2025年AI的产业重心从模型参数的比拼、算法的创新竞赛，逐步转向商业价值的深度挖掘与高效兑现，企业纷纷 将目光聚焦于如何让AI技术真正融入市场，转化为实际生产力与经济效益。 长视频内容创新方面，各个平台仍在积极推进内容精品战略，为观众提供更多优质剧集。 短剧则继续快速增长，中国网络视听协会最新测算的2025年微短剧行业全年产值已接近900亿元。 2025年第四季度正值中国传统的冬季旅游旺季，冰雪旅游成为带动旅游业增长的新引擎。 1.3互联网流量变化：移动设备达到12.99亿，用户时长相比三季度有所下滑 2.互联网细分市场发展分析 2.1AI篇：市场转向商业价值的深度挖掘与高效兑现 2.2视频篇：优质内容涌现，市场格局稳定 2.3电商篇：发展稳定，AI提升平台竞争力 2.4旅游篇：政策红利不断，冬季冰雪旅游火爆 展望2026年，互联网将继续稳定发展 ...

Group 1 - The Supreme People's Court ruled in favor of Yuzhu Technology in a patent infringement case against Luweimei, allowing Yuzhu to proceed with its IPO plans [2] - Tencent's Yuanbao team apologized for generating inappropriate content in user-created New Year greeting cards, addressing the issue and optimizing the model [2] - Hillhouse Capital plans to reduce its stake in Gree Electric Appliances by up to 112 million shares, with proceeds used to repay bank loans, marking its first reduction since becoming the largest shareholder [3] Group 2 - WeChat introduced new features for users to quickly locate images in chats and to receive photos and files face-to-face, enhancing user experience [4] - Yonyou Network clarified that its AI products are not intended to fully replace existing software but to enhance enterprise operations, focusing on knowledge workers [5] - Blue Arrow Aerospace announced plans to conduct recovery tests for its reusable rocket, Zhuque-3, in the second quarter of this year [6][7] Group 3 - Meizu's mobile business is reportedly ceasing operations, with plans to delist by March 2026, while its FlymeAuto division will operate independently [8] - Shunhao Co. announced that its subsidiary, Treadlight, plans to mass-produce satellites and establish a large-scale space data center between 2031 and 2035 [8] - The International Star Alliance updated its organizational structure to enhance technology implementation and scene promotion in smart home and terminal industries [9] Group 4 - Wingtech Technology is actively addressing the control rights issue of Anshi Semiconductor, focusing on maintaining business stability and communication with clients [10] - NVIDIA reported a revenue of $68.127 billion for Q4 of fiscal year 2026, a 73% year-over-year increase, with a net profit of $42.96 billion, reflecting a 94% growth [11] - Jiangsu Hongwei Technology announced a price increase for IGBT and MOSFET products due to rising manufacturing costs driven by fluctuations in core metal material prices [12][13] Group 5 - TrendForce predicts that the combined capital expenditure of eight major cloud service providers will exceed $710 billion by 2026, driven by AI server investments [14] - Manycore Tech Inc. updated its prospectus, reporting a revenue of 820 million yuan in 2025, with a gross margin of 82.2% [15] - Canaan Inc. completed the acquisition of several mining assets from Cipher Mining Technologies, shifting its strategy towards systematic upstream development in the U.S. power assets [16] Group 6 - Zhiyang Innovation plans to invest 300 million yuan in Lingming Photon Technology, acquiring a stake of approximately 9.09% to 10% [17]

Core Insights - The article discusses the transformation of business survival logic in the AI era, highlighting the decline of traditional competitive advantages established during the mobile internet age [2][6] - It emphasizes the shift from user loyalty and network effects to a focus on data accumulation and AI capabilities as the new competitive moat for companies [8][23] Group 1: AI Impact on Business Models - During the Spring Festival, over 130 million people experienced AI shopping, indicating a significant shift towards everyday AI applications [2] - The traditional "moats" of mobile internet, such as user habits and traffic logic, are being redefined in the face of AI advancements [2][6] - Companies must adapt to a new competitive landscape where user loyalty is minimal, and switching costs are low, as users prioritize functionality and performance over brand allegiance [6][7] Group 2: Data as a Competitive Advantage - The concept of "data compounding" is introduced as a crucial element for retaining users, where accumulated data about business processes and user habits becomes invaluable [9][23] - Three core dimensions of a data moat are identified: private context, interactive feedback loops, and industry-specific "dark knowledge" [12][14][16] - Companies that can structure and leverage internal data effectively will create barriers that competitors cannot easily overcome [14][24] Group 3: The Future of AI Applications - The article argues that the focus should shift from merely selecting AI models to designing systems that generate high-value feedback data during user interactions [20][23] - Successful B2B AI applications will integrate deeply into business processes, making data an essential part of operational flows [18][24] - As technology evolves, the true value lies in the unique data accumulated over time, which cannot be easily replicated by new models [23][24]