The AI Graveyard: Distressed Assets and the Correction of Venture Over-Leverage
AI-generated image · Bay Street Wire
As high-profile startups collapse and corporate AI initiatives fail, the industry is shifting toward a consolidation phase where surviving IP is absorbed by incumbents.
Q: What does the current rate of AI project failure suggest about the broader venture landscape?
A: The volume of shutdowns indicates a correction in how AI capabilities are valued and deployed. As TechCrunch first reported, the industry is seeing a growing "AI graveyard," where standalone products are struggling to maintain a reason to exist as larger platforms integrate similar features. This is evident in the case of Relay, a five-year-old workflow automation tool that shut down in August after OpenAI and Google built competing automation directly into their own ecosystems. Similarly, Character.ai filed for bankruptcy five years after its launch.
Q: Is this instability limited to small startups?
A: No. According to S&P Global Market Intelligence, roughly 42% of AI initiatives are abandoned by their corporate parents. These failures are attributed to a variety of factors, including technical hurdles, insufficient funding, difficulty scaling, competition, and weak user demand.
Q: How are the industry's largest players handling these failures internally?
A: Even dominant firms like OpenAI are pivoting away from standalone products toward consolidation. TechCrunch notes that OpenAI discontinued ChatGPT Atlas, an AI web browser, on August 9, and shut down Operator, an AI agent, folding their useful features into the main ChatGPT interface. OpenAI also shuttered its video-sharing platform, Sora, in March 2026 due to user-retention issues and high operating costs. Additionally, OpenAI attempted to launch a "super app" redesign on July 9 that combined "Work," "Codex," and "Chat" modes, but the company rolled back the change after users criticized the interface as cluttered.
Q: Are there examples of AI failures leading to distressed M&A and IP consolidation?
A: Yes. The collapse of AI hardware provides a clear example. TechCrunch reports that Humane raised $230 million from investors for its AI Pin, but the business shut down in February 2025 following performance struggles and battery fire risks. In a classic distressed exit, HP acquired most of Humane's assets for $116 million.
Q: What other operational or regulatory risks are contributing to this correction?
A: Execution delays and security concerns are creating significant financial and reputational liabilities. Apple faced a $250 million settlement related to the marketing of iPhone 16 AI capabilities after repeatedly delaying the release of its AI-powered Siri, which finally entered beta in July. Meanwhile, Microsoft's Recall feature, announced at the 2024 Build conference, faced a year-long delay due to privacy concerns. TechCrunch reports that despite a redesign, a cybersecurity researcher recently demonstrated a tool capable of extracting Recall's captured data, suggesting underlying security flaws persist.
Q: Are there any examples of products failing due to a shift in user behavior?
A: Yes. Notion Mail launched in April 2025 to automate inboxes, but the company discovered users preferred using separate AI agents for that purpose. Because of this, the service is set to close on September 22.