AI 阵亡名单：盘点那些走向关停或未达预期的明星项目与初创公司
英文标题：The AI graveyard: a running list of projects and startups that didn’t make it
随着生成式 AI 热潮逐步回归理性，以 Relay、Character.ai 等为代表的一批明星初创公司与前沿项目相继走向关停或申请破产。标普全球市场情报数据显示，企业发起的 AI 项目中有约 42% 最终被放弃，资金短缺、技术瓶颈、激烈竞争以及难以建立独立 PMF 是主要诱因。即便是 OpenAI、苹果、微软等头部科技巨头，在探索“超级应用”、AI 硬件与系统级 Agent 时也频繁遭遇产品回滚、延期与安全反弹。这座不断扩大的“AI 公墓”不仅记录了行业的失败教训，也向开发者与创业团队清晰揭示了单点 AI 工具在面对巨头生态整合与高昂运营成本时的生存困境。
3.1 KB
389 B
85.3 KB
13.6 KB
30.4 KB
28.4 KB
8.4 KB
23.2 KB
24.8 KB
38.3 KB
5.5 KB
53.0 KB
248 B
53.9 KB
中文
作为 Zapier 替代方案打造的 AI 工作流自动化工具 Relay 在今年 8 月关停。该产品允许用户使用 AI Agent 自动化处理邮件和任务工作流，但随着 OpenAI、Google 以及其他大型平台将类似的自动化能力直接内置到自有工具中，成立五年的 Relay 难以维持作为独立产品的生存理由。
不久前，聊天机器人平台 Character.ai 在上线仅五年后便申请破产。
除了 Character.ai 和 Relay 之外，许多 AI 赌注甚至未能在孵化它们的公司内部幸存。据标准普尔全球市场情报（S&P Global Market Intelligence）数据显示，企业母公司发起的 AI 项目中约有 42% 最终被放弃。原因多种多样：资金不足、技术挑战、竞争加剧、规模化困难，或者仅仅是用户需求疲软。
其结果是一座不断扩大的“AI 公墓”。以下盘点了一些最值得关注的已关停、转型或严重未达预期的 AI 产品、初创公司、发布项目与业务尝试。与其说这是一份失败清单，不如将其视作一份能为整个行业提供借鉴的历史记录。
OpenAI 的多次跌撞
图片来源： SeongJoon Cho/Bloomberg / Getty Images
即便是行业领头羊也无法免俗。作为全球估值最高的私营公司之一，OpenAI 同样经历过属于自己的阵痛与失误。最近的例子之一就是它试图将 ChatGPT 打造为更广泛的“超级应用”。7 月 9 日，OpenAI 对该应用进行了更新，将包括独立的“Chat”、“Codex”和“Work”模式在内的多种体验整合在一起，同时将用户熟悉的传统版本重命名为“ChatGPT Classic”。
这一改版迅速招致用户的批评，大家认为新界面既混乱又繁琐。OpenAI 在不久后便回滚了这一变更，重新带回了大家熟悉的 ChatGPT 界面。
这并非 OpenAI 唯一一次试图将所有功能收拢于同一屋檐下的尝试。该公司还关闭了多个独立应用，并将它们的功能直接合并到 ChatGPT 中——这一策略比前述改版运作得更有效。ChatGPT Atlas 是一款独立的 AI 网络浏览器，面世不到一年便于 8 月 9 日宣告停用，其最有价值的功能均被吸纳进了 ChatGPT。
该公司对 Operator 也采取了同样的处理方式。Operator 是一款能够代表用户浏览网页并执行任务的 AI Agent；同时，OpenAI 越来越倾向于将图像生成功能直接集成到 ChatGPT 中，从而削弱了 DALL-E 作为独立图像创作入口的角色。
我们同样不能忽略其视频分享平台 Sora，该平台因面临居高不下的运营成本与用户留存困境，已于 2026 年 3 月关停。
苹果的 Siri AI
图片来源： Artur Widak/NurPhoto / Getty Images
当苹果在 2024 年公布 Apple Intelligence 时，全面升级的 Siri 是其主打亮点之一。官方宣传承诺，Siri 将能够理解上下文、感知跨应用发生的操作，并高效完成任务。
然而，苹果以工程问题和 Bug等各种理由一再推迟新版 Siri 的上线。这一延期最终导致了一笔 2.5 亿美元的和解协议，以平息关于苹果在 iPhone 16 AI 功能宣传方面涉嫌误导的指控。
Siri AI 终于在今年 7 月的 iOS 27 公开测试版中现身，全新的 AI 版本本月率先向英语用户推送，对其他语言的支持预计随后跟进。
微软的 Recall
微软在 2024 年 Build 开发者大会上宣布的 Recall 功能，被宣传为 Windows PC 的 AI “照相式记忆”，它会定期对用户操作进行截屏，以便用户随后检索自己的数字化历史记录。
不出所料，该声明立即引发了围绕隐私和安全的强烈反弹。批评者指出，Recall 可能会形成一个包含密码、私人信息、财务细节和其他敏感活动的可搜索归档库。微软将该功能推迟了近一年，以便有时间重新审视其安全与隐私保护机制。
即便经过重新设计，争议仍未平息。一位网络安全研究人员最近构建了一个工具，能够提取并展示 Recall 捕获的数据，再次引发了人们对微软是否真正解决了底层安全问题的质疑。
Notion Mail
图片来源： Notion
Notion Mail 于 2025 年 4 月上线，是一款侧重 AI 的电子邮件产品，旨在帮助用户整理和自动化收件箱。然而，该公司发现用户更倾向于使用独立的 AI Agent 来管理邮箱。因此，Notion Mail 计划于 9 月 22 日正式停运。
Humane AI Pin
图片来源： Brian Heater
Humane AI Pin 大概是 AI 硬件领域最广为人知的失败案例之一。该产品旨在通过可穿戴设备而非传统智能手机提供 AI 功能，Humane 从投资者手中筹集了 2.3 亿美元并获得了极大关注。
然而，该产品在性能表现上面临严峻问题；当 Humane 因电池起火隐患而警告用户停用充电盒时，情况进一步恶化。Humane 随后于 2025 年 2 月关停了 AI Pin 业务，公司大部分资产被惠普以 1.16 亿美元收购。
Rabbit R1
图片来源： Rabbit
Rabbit R1 是另一款被高度炒作的 AI 硬件设备。这款在 2024 年 1 月 CES 上亮相的 AI 伴侣设备，旨在代表所有者执行任务。Rabbit 曾表示其在发布后不久便售出了 10 万台。
然而，强劲的初期销量并不等同于过硬的产品力。早期评测将该设备描述为半成品，性能不可靠且实用的应用集成非常有限。
不过该公司并未放弃。它持续对 R1 进行更新，将其重新定位为能够执行 Agent 任务的电脑控制器。此外，它最近还公布了一个名为“Project Cyberdeck”的新硬件项目，旨在打造一款专为氛围编码（vibe-coding）设计的便携式设备。
Huxe
图片来源： Huxe
Huxe 是一款 AI 音频应用，由前 Google NotebookLM 开发者打造，能够将书面文本转化为对话式的播客风格音频。
该公司于 2026 年 5 月关停，部分原因是体量更大的平台开始向其现有的庞大用户群提供类似体验。例如，Spotify 一直在播客中拓展自研的 AI 驱动音频工具。
Yupp
图片来源： Yupp
Yupp 是一款有趣的 AI 产品：一个允许用户并排对比数百个 AI 模型回答的免费实验平台。用户还可以对回答进行投票并通过该平台赚取加密货币。
在巅峰时期，该平台允许用户测试超过 800 个模型，其中包括来自 OpenAI、Google 和 Anthropic 等公司的模型。但据其创始人透露，Yupp 从未获得足以生存的强劲产品市场契合度（PMF）。该平台于 2026 年 3 月关停。
Figgs AI
Figgs AI 于 2023 年至 2024 年期间运营，允许用户创建自定义 AI 角色并与之交互，用于角色扮演和故事创作。据报道该平台吸引了超过 100 万用户，但其开发者表示维持免费服务的成本过高难以维系，最终关停了服务。
Character.ai
由前 Google 研究员 Noam Shazeer 和 Daniel De Freitas 于 2021 年创立的 Character.ai，凭借允许用户与自定义机器人对话（许多机器人以名人和虚构角色为原型）而成为风靡一时的消费级 AI 应用。它在 2023 年以 10 亿美元估值完成了 1.5 亿美元融资。次年，Google 据悉斥资 27 亿美元获得了其技术授权并将创始人召回麾下，留下这家初创公司在没有创始人的情况下继续运营。
此后麻烦接踵而至。多家家庭发起诉讼，指控该聊天机器人导致青少年自杀和自残；迪士尼也因其未经授权使用角色而发出了停火函（Cease-and-desist letter）。Character.ai 于去年年底禁止 18 岁以下未成年人进行开放式聊天，并于 2026 年初与 Google 一同就多起诉讼达成和解。但这仍不足以力挽狂澜。10 月初，该公司以法律、监管和运营阻力为由申请破产。讽刺的是，迪士尼如今成为了首位竞标者，出资 2000 万美元寻求获得其部分知识产权的许可。在破产重组流程期间，该应用仍在正常运行，预计出售流程将在 11 月下旬或 12 月初完成。
English / 原文
Relay, an AI-powered workflow automation tool built as an alternative to Zapier, shut down entirely in August. The product let users automate email and task workflows using AI agents, but as OpenAI, Google, and other larger platforms built similar automation directly into their own tools, five-year-old Relay struggled to hold on to a reason to exist as a stand-alone product.
More recently, chatbot platform Character.ai filed for bankruptcy just five years after launching.
Aside from Character.ai and Relay, plenty of AI bets never even survive inside the companies that build them. According to S&P Global Market Intelligence, about 42% of AI initiatives are ultimately abandoned by their corporate parents. The reasons vary: insufficient funding, technical challenges, competition, difficulty scaling, or just weak user demand.
The result is a growing AI graveyard. Below is a look at some of the most notable AI products, startups, launches, and bets that have shut down, pivoted, or significantly missed expectations. Think of it less as a scorecard of flops than as a record of what these attempts can (hopefully) teach the rest of the industry.
OpenAI has taken several stumbles
Image Credits: SeongJoon Cho/Bloomberg / Getty Images
Even the biggest players aren’t immune. OpenAI, one of the most valuable private companies in the world, has had its own share of casualties and missteps. One of the more recent examples was its attempt to turn ChatGPT into a broader “super app.” On July 9, OpenAI updated the app to combine several experiences, including separate “Chat,” “Codex,” and “Work” modes, while renaming the traditional, familiar version “ChatGPT Classic.”
The redesign quickly drew criticism from users who found the new interface confusing and cluttered, and OpenAI rolled back the change soon after, bringing the familiar ChatGPT interface back.
That wasn’t OpenAI’s only attempt to pull everything under one roof. The company has also shut down several stand-alone apps and folded their features directly into ChatGPT — a strategy that’s worked out better than the redesign did. ChatGPT Atlas, a stand-alone AI-powered web browser, lasted less than a year before being discontinued on August 9, with its most useful features absorbed into ChatGPT.
The company did the same thing with Operator, an AI agent that could browse the web and complete tasks on a user’s behalf, and OpenAI has increasingly built image generation directly into ChatGPT, shrinking the role of DALL-E as its own separate destination for creating images.
We also can’t forget its video-sharing platform, Sora, which struggled with high operating costs and user-retention challenges and shut down in March 2026.
Apple’s Siri AI
Image Credits: Artur Widak/NurPhoto / Getty Images
When Apple unveiled Apple Intelligence in 2024, an improved Siri was one of the highlighted features. The pitch promised a Siri that could understand context, know what was happening across apps, and get tasks done effectively.
However, Apple repeatedly pushed back the release of the new Siri, citing various reasons like engineering issues and bugs. This delay eventually contributed to a $250 million settlement over claims regarding how Apple marketed the AI capabilities of the iPhone 16.
Siri AI finally appeared in the iOS 27 beta this past July, with the new AI-powered version rolling out to English-language users first this month; support for other languages is expected to follow.
Microsoft’s Recall
Announced at its 2024 Build developer conference, Microsoft’s Recall feature was pitched as an AI-powered “photographic memory” for Windows PCs, periodically taking screenshots of what a user was doing so they could later search back through their own digital history.
Predictably, the announcement triggered an immediate backlash over privacy and security. Critics argued that Recall could create a searchable archive of highly sensitive information, including passwords, private messages, financial details, and other activity. Microsoft delayed the feature for nearly a year, giving the company time to rethink its security and privacy protections.
Even after a redesign, the controversy hasn’t gone away. A cybersecurity researcher recently built a tool that could extract and display data that Recall had captured, reviving questions about whether Microsoft has actually fixed the underlying problem.
Notion Mail
Image Credits: Notion
Notion Mail launched in April 2025 as an AI-focused email product designed to help users organize and automate their inboxes. However, the company noticed its users were using separate AI agents to handle their email inboxes instead. As a result, Notion Mail is scheduled to shut down on September 22.
Humane AI Pin
Image Credits: Brian Heater
The Humane AI Pin is probably one of the most well-known failures in AI hardware. Designed to provide AI features through a wearable device rather than a traditional smartphone, Humane raised $230 million from investors and garnered significant attention.
But the product struggled mightily with performance, and the situation worsened when Humane warned customers to stop using its charging case because of a potential battery fire risk. Humane then shut down its AI Pin business in February 2025, with most of the company’s assets acquired by HP for $116 million.
Rabbit R1
Image Credits: Rabbit
The Rabbit R1 was another heavily hyped AI device. Unveiled at CES in January 2024, the AI companion was designed to perform tasks on behalf of its owner. Rabbit said it had sold 100,000 units shortly after launch.
However, strong initial sales didn’t mean it was a strong product. Early reviews described the device as unfinished, with unreliable performance and a limited number of useful integrations.
The company hasn’t given up, though. It has continued updating the R1, positioning it as a computer controller capable of performing agentic tasks. It also recently announced a new hardware project, dubbed “Project Cyberdeck,” aimed at creating a portable device built for vibe-coding.
Huxe
Image Credits: Huxe
Huxe was an AI audio app, built by former developers of Google’s NotebookLM, that turned written information into conversational, podcast-style audio.
The company shut down in May 2026, partly as larger platforms began offering similar experiences to large existing audiences. For instance, Spotify has been expanding its own AI-powered audio tools.
Yupp
Image Credits: Yupp
Yupp was an interesting AI product: a free playground that allowed users to compare responses from hundreds of AI models side by side. Users could also vote on responses and earn cryptocurrency through the platform.
At its peak, the platform let users test more than 800 models, including from companies such as OpenAI, Google, and Anthropic. But according to its founders, Yupp never achieved strong enough product-market fit to survive. The platform shut down in March 2026.
Figgs AI
Figgs AI operated from 2023 to 2024, letting users create and interact with customizable AI characters for role-play and storytelling. The platform reportedly drew more than 1 million users, but its developers said keeping the service free became too expensive to sustain, and they shut it down.
Character.ai
Founded in 2021 by former Google researchers Noam Shazeer and Daniel De Freitas, Character.ai became a buzzed-about consumer AI app by letting people chat with customizable bots, many of them modeled on celebrities and fictional characters. It raised $150 million at a $1 billion valuation in 2023. Then, the next year, Google paid a reported $2.7 billion to license its technology and bring its founders back into its orbit, leaving the startup to carry on without them.
The trouble piled up after that. Families sued, alleging the chatbots contributed to teen suicides and self-harm, and Disney sent a cease-and-desist letter over unauthorized use of its characters. Character.ai barred under-18 users from open-ended chats late last year and agreed to settle several of the suits, along with Google, in early 2026. But it wasn’t enough. The company filed for bankruptcy in early October, citing legal, regulatory, and operational headwinds. In an ironic twist, Disney is now the opening bidder, offering $20 million for a license to some of its intellectual property. The app is still running during the process, and the sale is expected to wrap up by late November or early December.