AI industry faces widespread project failures as competition and technical barriers mount
A growing list of AI products and startups—from OpenAI's ChatGPT Atlas to Humane's AI Pin—have shut down, been abandoned, or significantly underperformed, revealing structural challenges across the sector.
What happened
Multiple high-profile AI projects have failed or been discontinued in recent months. Relay, a five-year-old AI workflow automation tool, shut down entirely in August as larger platforms like OpenAI and Google built similar features directly into their own products. Character.ai filed for bankruptcy five years after launch. OpenAI attempted to redesign ChatGPT into a broader 'super app' in July but rolled back the change after user backlash. OpenAI has also shut down several standalone apps including ChatGPT Atlas (a web browser discontinued August 9) and Operator (an AI agent). Sora, OpenAI's video-sharing platform, shut down in March 2026 due to high operating costs and user-retention challenges. Notion Mail, launched in April 2025 as an AI-focused email product, is scheduled to shut down on September 22 after users opted for separate AI agents instead. Hardware devices have also struggled: Humane AI Pin, which raised $230 million, shut down in February 2025 after performance issues and a battery fire risk warning; HP acquired most of the company's assets for $116 million. Rabbit R1 was unveiled at CES in January 2024 and sold 100,000 units shortly after launch, but early reviews described it as unfinished with unreliable performance, though the company continues updating it. According to S&P Global Market Intelligence, approximately 42% of AI initiatives are ultimately abandoned by their corporate parents.
Context
The pattern of failures reflects fundamental challenges facing the AI industry: larger incumbents can absorb similar functionality into existing platforms, making standalone products difficult to sustain; technical barriers remain high, with execution problems plaguing even well-funded ventures; and user demand for specialized AI tools may not justify standalone products when alternatives exist. Apple unveiled Apple Intelligence in 2024 with an improved Siri as a highlighted feature, but repeatedly pushed back the release, citing engineering issues and bugs. An improved Siri finally appeared in iOS 27 beta in July, contributing to a $250 million settlement over Apple's marketing claims regarding AI capabilities of the iPhone 16. Microsoft announced Recall at its 2024 Build conference as an AI 'photographic memory' for Windows, but the announcement triggered immediate backlash over privacy and security concerns. Microsoft delayed the feature for nearly a year to rethink its security and privacy protections, but cybersecurity researchers have since demonstrated that data extraction from Recall remains possible despite redesigns, suggesting that fundamental trust and security issues can persist even after attempts to address them. These failures indicate a divergence between AI hype and product-market fit, with investor attention and initial interest not translating to sustainable