# AI时代首个互联网拐点出现：机器人阅读访问量首次超越人类 - ID: 75 - URL: https://www.stoyard.com/viewpoint/75 - English URL: https://www.stoyard.com/en/viewpoint/75 - Category: AI 探索 - English Category: AI exploration - English Title: The first Internet inflection point in the AI era appears: robot reading visits surpass human beings for the first time ## 中文主体 > 2026年，一个可能被未来互联网历史反复提及的时刻出现了。根据全球互联网基础设施公司 Cloudflare 最新公布的数据： 2026年，一个可能被未来互联网历史反复提及的时刻出现了。根据全球互联网基础设施公司 Cloudflare 最新公布的数据： ### ### 全球网页HTTP请求中，机器人与AI系统产生的流量占比已达到57.5%，而人类用户仅占42.5%。  这意味着：自互联网诞生以来，机器首次成为互联网最大的“访问者”。Cloudflare CEO Matthew Prince 表示，原本预计这一拐点将在2027年出现，但实际发生时间比预测提前了一年。(Tom's Hardware)。对于整个互联网产业来说，这不仅仅是一组流量数据。 ### 互联网的主要读者，正在从人类变成机器。互联网正在被机器重新定义 过去二十年，我们一直生活在“人类互联网”时代。其核心逻辑都只有一个：让人看到。但今天，这个前提正在发生改变。这也是互联网历史上第一次，机器访问量超过人类访问量。(Tom's Hardware)与此同时，网络安全公司 Human Security 发布的《State of AI Traffic》报告显示： - 2025年AI相关流量增长187% - Agent（智能体）流量增长超过8000% - AI驱动流量成为互联网增长最快的流量类别 (纽约邮报) 如果说2023年是AI聊天时代。那么2026年，已经正式进入AI代理（Agent）时代。 --- ### 过去的爬虫，只是搬运工，今天的AI，已经成为决策者 很多人会说：“互联网一直都有机器人。”确实如此。Googlebot、百度蜘蛛、各种搜索引擎爬虫已经存在二十多年。 但过去的机器人只负责： - 抓取网页 - 建立索引 - 服务人类搜索 它们本身不做判断。而今天的AI Agent不同。 它可以： ✓ 阅读网页 ✓ 提取信息 ✓ 比较产品 ✓ 分析参数 ✓ 总结优缺点 ✓ 直接给出决策建议 Cloudflare将这种新型流量称为：Agentic Traffic（智能代理流量）这是推动机器流量首次超越人类流量的核心原因。(Tom's Hardware) --- ### 第一个被改变的行业：广告业 互联网过去20年的商业模式，本质上都是：流量 → 点击 → 转化。广告行业能够持续增长的前提是：人会被广告影响。 但AI不会： - 被品牌故事感动 - 被营销话术说服 - 被视觉包装影响 AI关注的： - 数据 - 参数 - 评价 - 客户案例 - 技术指标 Cloudflare CEO Matthew Prince近期公开表示： AI正在挑战互联网长期以来依赖的广告商业模式。因为机器人不会点击广告。更不会被广告说服。(纽约邮报)。这意味着：未来企业最大的竞争可能不是SEO。而是：AEO（AI Engine Optimization）即：如何让AI优先推荐你。 --- ### ### 第二个被改变的行业：信息中介 过去很多商业模式依赖信息不对称。 例如： - 渠道商 - 信息聚合平台 - 参数搬运内容 - 比价平台 这些模式都有一个共同点：掌握别人不知道的信息。但AI最擅长的事情恰恰是：消灭信息差。它可以同时阅读：100个网页，1000条评价，数十份报告。然后在数秒内完成分析。对于消费者来说：以前需要3小时完成的研究工作。现在可能只需要一句话：“帮我推荐最适合家庭使用的空气净化器。”AI已经完成了背后的所有比对工作。 --- ### 第三个被改变的行业：企业官网 这也是大多数企业尚未意识到的变化。过去的网站设计逻辑是：让用户看懂。未来的网站设计逻辑将变成：让AI读懂。因为越来越多用户将不再直接访问网站。 而是先询问AI： - 推荐一家设计公司 - 推荐一家酒店集团供应商 - 推荐一家医疗器械企业 - 推荐一家技术服务商 然后AI替用户完成筛选。企业官网正在从：品牌展示工具。变成：AI时代的数据基础设施。 --- ### ### 一个更值得关注的数据 Cloudflare在2025年下半年仅5个月时间里： 累计拦截了超过4160亿次AI爬虫抓取请求。 平均每天约27亿次。 AI正在以前所未有的速度读取互联网内容。(Tom's Hardware) 与此同时，Cloudflare 2025年度报告显示： 全球互联网流量增长19%。 其中相当大比例来自机器人和AI系统。(TechRadar) 互联网增长还在继续。 但增长主体已经发生变化。 --- ### ### 互联网正在进入“机器优先”时代 过去二十年：企业研究用户。 未来十年：企业不仅要研究用户。还要研究AI。 因为未来的流量路径将变成：用户 → AI → 企业 而不是：用户 → 搜索引擎 → 企业 这是整个互联网底层分发逻辑最大的变化。 --- 2000年左右，我们进入搜索引擎时代。 2010年左右，我们进入移动互联网时代。 2020年以后，我们进入社交媒体时代。 而2026年，随着机器人访问量首次超过人类访问量，一个新的时代正式开启：AI代理时代。 未来的网站、内容、产品和品牌，可能不再首先面对人类。而是首先面对机器。 谁能被AI理解。 谁就能被AI推荐。 谁能被AI推荐。 谁就将获得下一轮流量红利。 ## English Content > In 2026, a moment that may be repeatedly referenced in the annals of future internet history emerged. According to the latest data released by Cloudflare, a global internet infrastructure company: In 2026, a moment that may be repeatedly referenced in the annals of future internet history emerged. According to the latest data released by Cloudflare, a global internet infrastructure company: ### ### Among global web HTTP requests, the share of traffic generated by bots and AI systems has reached 57.5%, while human users account for only 42.5%.  This means that, since the birth of the Internet, machines have, for the first time, become its largest “visitors.” Cloudflare CEO Matthew Prince stated that the company had originally projected this inflection point would occur in 2027, but it actually arrived a year earlier than expected. (Tom's Hardware). For the entire internet industry, this is far more than just a set of traffic metrics. ### The primary audience of the internet is shifting from humans to machines. The internet is being redefined by machines. Over the past two decades, we have been living in the era of the “Human Internet.” Its core logic is all the same: to get people to see it. But today, this premise is undergoing a change.This also marks the first time in internet history that machine traffic has surpassed human traffic. (Tom's Hardware) Meanwhile, the State of AI Traffic report released by cybersecurity firm Human Security shows: -AI-related traffic will grow by 187% in 2025 -Agent traffic has grown by more than 8,000% -AI-driven traffic has become the fastest-growing traffic category on the internet (The New York Post) If 2023 is the era of AI chat. So by 2026, we will have officially entered the AI agent era. --- ### The reptiles of the past were only porters, but today's AI have become decision makers. Many people will say, "The Internet has always had robots." Indeed. Googlebot, Baidu Spider, and various search engine crawlers have been around for more than two decades. But in the past, robots were only responsible for: -Fetch web pages -Create an index -Serving human search They do not make judgments on their own. But today’s AI Agent is different. It can: ✓ Read web page ✓ Extract information ✓ Compare products ✓ Analysis parameters ✓ Summarize the pros and cons ✓ Provide decision recommendations directly Cloudflare call this new type of traffic: Agentic Traffic (intelligent agent traffic) This is the core reason that drives machine traffic to surpass human traffic for the first time. (Tom's Hardware) --- ### The first industry to be transformed: the advertising industry The business model of the Internet in the past 20 years is essentially: traffic → click → transformation. The premise that the advertising industry can continue to grow is that people will be affected by advertising. But AI won’t: -Moved by the brand’s story -Persuaded by marketing rhetoric -Influenced by visual packaging AI's focus on: -Data -Parameter -Review -Customer Case Studies -Technical indicators Cloudflare CEO Matthew Prince recently stated publicly: AI is challenging the advertising business model that the internet has long relied on. Because robots don’t click on ads. And I’m even less likely to be swayed by advertisements. (The New York Post). This means that in the future, the greatest competitive challenge for businesses may not be SEO. Rather, it is: AEO (AI Engine Optimization), which refers to how to make AI prioritize recommending you. --- ### ### The second industry to be transformed: information intermediation In the past, many business models relied on information asymmetry. For example: -Channel partner -Information aggregation platform -Parameter transfer content -Price comparison platform These patterns all have one thing in common: having information that no one else knows. But what AI do best is precisely this: wipe out poor information. It can read 100 web pages, 1000 reviews, dozens of reports simultaneously. Then complete the analysis within a few seconds. For consumers: research that used to take 3 hours to complete. Now it may only need one sentence: "help me recommend the most suitable air purifier for family use." AI has already completed all the matching work behind the scenes. --- ### The third industry to be transformed: corporate websites This is also a change that most companies have yet to recognize. In the past, the website design logic is: let users understand. The future of website design logic will become: let AI read. Because an increasing number of users will no longer access websites directly. Instead, first ask the AI: -Recommend a design company -Recommending a hotel group supplier -Recommending a medical device company -Recommend a technology service provider Then the AI completes the screening for the user. Enterprise official website is from: brand display tools. Becomes: The data infrastructure of the AI era. --- ### ### A more noteworthy piece of data In the second half of 2025, over a period of just five months: A total of over 416 billion AI bot scraping requests have been intercepted. An average of approximately 2.7 billion times per day. AI is reading internet content at an unprecedented speed. (Tom's Hardware) Meanwhile, Cloudflare’s 2025 annual report indicates: Global internet traffic increased by 19%. A considerable proportion of it comes from robots and AI systems. (TechRadar) Internet growth is still ongoing. However, the main driver of growth has already shifted. --- ### ### The internet is entering the “machine-first” era. Over the past twenty years: companies have been studying users. Over the next decade, companies must go beyond studying users. We also need to study AI. Because the future traffic pathway will become: user → AI → enterprise Instead of: User → Search Engine → Business This is the most significant change to the underlying distribution logic of the entire internet. --- Around the year 2000, we entered the era of search engines. Around 2010, we entered the era of the mobile internet. Since 2020, we have entered the era of social media. And in 2026, as robot traffic surpasses human traffic for the first time, a new era officially begins: the AI agent era. In the future, websites, content, products, and brands may no longer be designed primarily with humans in mind. Instead, we first face the machine. Who can be understood by AI. Whoever it is will be recommended by AI. Who can be recommended by AI. Whoever does will capture the next wave of traffic红利.