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

A Reckoning for AI Startups

A growing number of AI products are being shuttered, highlighting a harsh new reality for startups and corporate projects alike. The recent bankruptcy filing of chatbot platform Character.ai and the complete shutdown of workflow automation tool Relay signal a difficult market where even well-funded ventures are struggling to survive. This trend isn't isolated; it underscores the immense pressure startups face when competing against major platforms like OpenAI and Google, which are increasingly integrating similar AI features directly into their own ecosystems, leaving little room for standalone applications.

The reasons for these failures span technical, financial, and market-driven challenges. AI hardware has seen notable flameouts, with the Humane AI Pin ceasing operations after struggling with poor performance and safety issues, while the Rabbit R1 received criticism for being unfinished despite strong initial sales. In the software space, products like Relay, Huxe, and Notion Mail were ultimately cannibalized by larger competitors offering similar functionality. According to S&P Global Market Intelligence, roughly 42% of corporate AI initiatives are eventually abandoned due to a variety of factors, including:

• Intense competition from incumbent platforms
• Technical challenges and difficulty scaling
• Privacy and security backlash, as seen with Microsoft's Recall
• Weak user demand or a failure to find product-market fit

The Great Consolidation

The impact of these closures points toward a broader market consolidation. Even industry leader OpenAI has stumbled, rolling back a confusing ChatGPT redesign and discontinuing standalone apps like the Atlas browser to fold their features into its central product. This strategy, while effective for incumbents, creates a hostile environment for startups built around a single feature. The closures serve as a record of what doesn't work, teaching the industry that a compelling AI function is no longer enough to build a sustainable business when it can be replicated as a feature by a platform with massive distribution.

The recent AI graveyard is less an indictment of innovation and more a clear signal that standalone AI features struggle to survive. Startups are learning a hard lesson: unless a product offers a defensible, platform-level advantage, it risks being absorbed or rendered obsolete by incumbents like OpenAI and Google who can bundle similar capabilities at scale.