The AI graveyard: A running list of projects and startups that didn’t make it

Relay, an AI-driven workflow automation platform positioned as a Zapier alternative, ceased operations on Monday. The service enabled users to automate email and task workflows via AI agents, but as OpenAI, Google, and other major platforms integrated similar automation directly into their ecosystems, the five-year-old Relay struggled to justify its existence as a standalone product.

Relay, a startup outpaced by larger platforms, represents one facet of the current industry landscape. However, many AI initiatives never survive within the companies that develop them. According to S&P Global Market Intelligence, approximately 42% of corporate AI initiatives are ultimately abandoned. Reasons for failure vary, including insufficient funding, technical hurdles, competitive pressure, scaling difficulties, or weak user demand.

This has resulted in a growing AI graveyard. The following overview highlights notable AI products, startups, launches, and initiatives that have shut down, pivoted, or significantly missed expectations. This list serves less as a scorecard of failures and more as a record of lessons that these attempts can hopefully impart to the rest of the industry.

OpenAI has encountered several setbacks

Photographer: SeongJoon Cho/Bloomberg via Getty ImagesImage Credits:SeongJoon Cho/Bloomberg / Getty Images

Even industry giants are not immune to missteps. OpenAI, one of the world’s most valuable private companies, has experienced its own share of casualties. A recent example was its attempt to transform ChatGPT into a comprehensive “super app.” On July 9, OpenAI updated the application to consolidate several experiences, including distinct “Chat,” “Codex,” and “Work” modes, while renaming the traditional version “ChatGPT Classic.”

The redesign faced immediate user criticism for being confusing and cluttered, prompting OpenAI to reverse the change and restore the familiar ChatGPT interface shortly thereafter.

This was not OpenAI’s only effort to consolidate features under one umbrella. The company has also discontinued several standalone applications, integrating their functionalities directly into ChatGPT—a strategy that has proven more successful than the interface redesign. ChatGPT Atlas, a standalone AI-powered web browser, was discontinued on August 9, less than a year after launch, with its most valuable features absorbed into ChatGPT.

The company applied a similar approach to Operator, an AI agent capable of browsing the web and executing tasks on behalf of users. Additionally, OpenAI has increasingly integrated image generation directly into ChatGPT, diminishing the standalone role of DALL-E as a dedicated platform for image creation.

Furthermore, its video-sharing platform, Sora, struggled with high operating costs and user retention issues, leading to its shutdown in March 2026.

Apple’s Siri AI

Photo by Artur Widak/NurPhoto via Getty ImagesImage Credits:Artur Widak/NurPhoto / Getty Images

When Apple introduced Apple Intelligence in 2024, an enhanced Siri was a key feature. The promise was a Siri capable of understanding context, tracking activity across apps, and executing tasks effectively.

However, Apple repeatedly delayed the new Siri’s release, citing engineering issues and bugs. This delay ultimately contributed to a $250 million settlement regarding claims about how Apple marketed the AI capabilities of the iPhone 16.

Siri AI finally appeared in the iOS 27 beta last 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 marketed as an AI-powered “photographic memory” for Windows PCs. It periodically captured screenshots of user activity, allowing users to search through their digital history.

Predictably, the announcement triggered 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 activities. Microsoft delayed the feature for nearly a year, allowing time to rethink its security and privacy protections.

Even after a redesign, controversy persists. A cybersecurity researcher recently developed a tool capable of extracting and displaying data captured by Recall, reviving questions about whether Microsoft has truly resolved the underlying security issues.

Notion Mail

Notion Mail uses AI to help users focus their inbox. 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 observed that users were employing separate AI agents to manage their email 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 likely one of the most prominent failures in AI hardware. Designed to deliver AI features through a wearable device rather than a traditional smartphone, Humane raised $230 million from investors and garnered significant attention.

However, the product struggled significantly with performance. The situation worsened when Humane warned customers to stop using its charging case due to a potential battery fire risk. Humane subsequently 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 claimed to have sold 100,000 units shortly after launch.

However, strong initial sales did not equate to a strong product. Early reviews described the device as unfinished, citing unreliable performance and a limited number of useful integrations.

The company has not 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 converted written information into conversational, podcast-style audio.

The company shut down in May 2026, partly because larger platforms began offering similar experiences to their vast 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 allowed users to test more than 800 models, including those from OpenAI, Google, and Anthropic. However, 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, allowing users to create and interact with customizable AI characters for roleplay and storytelling. The platform reportedly attracted more than 1 million users, but its developers stated that keeping the service free became too expensive to sustain, leading to its shutdown.