What changed
Last month, chatbot platform Character.ai filed for bankruptcy just five years after launching. More recently, privacy-focused AI assistant Ollie announced its shutdown a few weeks ago.
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.
What this means in practice
Character.ai, founded in 2021 by former Google researchers Noam Shazeer and Daniel De Freitas, was a consumer AI app that let people chat with customizable bots. It raised $150 million at a $1 billion valuation in 2023.
However, the company faced legal and regulatory challenges, including lawsuits from families alleging the chatbots contributed to teen suicides and self-harm. Character.ai also faced issues with Disney over unauthorized use of its characters.
The company filed for bankruptcy in early October, citing legal, regulatory, and operational headwinds.
Ollie, launched in June 2026 as one of the first mainstream apps to achieve SOC 2 compliance, aimed to bring together family information across various platforms. However, it struggled to stand out against competitors like Muse and Instinct, leading to its shutdown on November 6.
Limitations and availability
The supplied announcement does not specify the exact number of AI initiatives that have failed or the total number of AI projects currently in development. The failure rate of 42% is based on data from S&P Global Market Intelligence.
The reasons for failure include insufficient funding, technical challenges, competition, difficulty scaling, and weak user demand. The article highlights several notable examples, including OpenAI's attempts to integrate various AI features into ChatGPT, which led to mixed results and some shutdowns.
Microsoft's Recall feature, which faced significant privacy concerns, is another example of the challenges in the AI space. The article also mentions the failure of hardware projects like the Humane AI Pin and Rabbit R1, which struggled with performance and reliability issues.
