AI graveyard grows: Character.ai, Ollie, and others shut down

Character.ai’s bankruptcy follows legal and leadership troubles

Character.ai, the AI chatbot startup founded by former Google researchers Noam Shazeer and Daniel De Freitas, has filed for bankruptcy in October 2026. The company had raised $150 million at a $1 billion valuation in 2023, but legal challenges, regulatory pressures, and operational struggles overwhelmed it. Google had earlier licensed its technology for $2.7 billion and rehired its founders, leaving the startup to operate without its original leadership. Disney has since emerged as the opening bidder for parts of its intellectual property, offering $20 million for a limited license. The bankruptcy process is expected to conclude by December 2026.

Ollie shuts down after failing to stand out

Ollie, a privacy-focused AI assistant for families, shut down on November 6, 2026. The startup, launched in June 2026, had achieved SOC 2 compliance but struggled to differentiate itself in a crowded market dominated by larger players like Muse and Instinct. CEO Bill Lennon cited challenges in growth, visibility, and fundraising as key reasons for the shutdown.

Nearly half of AI projects fail

Nearly 42% of corporate AI initiatives are abandoned before reaching maturity, according to S&P Global Market Intelligence. The reasons range from insufficient funding and technical hurdles to competition and weak user demand. This trend spans startups and established companies, with OpenAI, Apple, and Microsoft all withdrawing or delaying high-profile projects in 2026.

OpenAI consolidates as standalone projects fail

OpenAI has discontinued multiple standalone projects this year, including ChatGPT Atlas, Operator, and Sora. The company is consolidating features into its flagship ChatGPT platform, a strategy that has faced mixed reactions from users. Atlas, an AI-powered web browser, was shut down in August 2026 after less than a year, while Operator, an AI agent for task automation, met a similar fate. Sora, OpenAI’s video-sharing platform, closed in March 2026 due to high operating costs and poor user retention.

Apple’s delayed Siri AI leads to legal settlement

Apple’s revamped Siri AI, announced as part of Apple Intelligence in 2024, faced repeated delays and engineering challenges. The company finally rolled out the updated version in October 2026, but only for English-language users. The delays contributed to a $250 million settlement over claims that Apple overstated the AI capabilities of the iPhone 16.

Microsoft Recall delayed amid privacy backlash

Microsoft’s Recall feature, introduced at its 2024 Build conference, was delayed for nearly a year due to privacy concerns. The tool, designed to take periodic screenshots of user activity for later search, faced backlash over potential security risks. Its future remains uncertain as Microsoft revisits its approach.

Hardware and niche AI startups struggle to survive

Humane AI Pin, a well-funded hardware startup, shut down in February 2025 after raising $230 million. HP acquired most of its assets for $116 million. Rabbit R1, another hardware play, sold 100,000 units but was criticized for unreliable performance and an unfinished product experience. Smaller startups like Yupp and Figgs AI also shut down in 2026, citing weak product-market fit and funding challenges.

Key takeaways

• AI startups and projects are failing at an accelerating pace, with nearly half of corporate AI initiatives abandoned due to funding shortages, technical challenges, or weak demand. This trend matters for founders and operators because it reveals the high bar for product-market fit in AI—even well-funded teams struggle to scale or differentiate.
• OpenAI’s shutdowns of standalone projects like Operator and Sora, and its pivot to consolidating features into ChatGPT, signal a shift toward integration over fragmentation. Startups building niche AI tools may need to prove they can coexist with or outperform platform-level solutions.
• Privacy-focused AI assistants like Ollie and hardware like Humane AI Pin failed despite strong positioning, showing that compliance and innovation alone aren’t enough without a clear path to user growth and capital. Founders in regulated spaces must plan for longer runways.
• Delaying or overpromising AI features—like Apple’s revised Siri—can lead to legal and financial repercussions, including settlements and reputational damage. Transparency about capabilities and timelines is becoming critical as regulators and users grow more skeptical.

Related on Lazyfounder

• OpenAI
• Apple
• Google
• Microsoft
• Disney
• More ai news
• Meta’s Muse AI calls businesses—many just hang up