Goliath Super Intelligence
IndustryOctober 25, 20262 min read

Industry faces mounting wave of AI project shutdowns and pivots

From workflow tools to hardware wearables, dozens of AI ventures have folded or been restructured, highlighting funding gaps, technical hurdles, and regulatory pressures.

A growing number of artificial-intelligence experiments are ending prematurely. Data from S&P Global Market Intelligence suggests that roughly four out of ten corporate AI projects are eventually discontinued. The causes range from insufficient capital and technical roadblocks to stiff competition and weak user demand, creating a widening AI graveyard that industry observers are tracking closely.

Relay, an AI-driven workflow automation service positioned as a Zapier alternative, ceased operations in August after five years. Its core function of automating email and task flows was eclipsed as larger platforms such as OpenAI and Google embedded similar capabilities directly. Meanwhile, chatbot provider Character.ai entered bankruptcy in October, five years after its launch, despite earlier high valuations.

OpenAI’s attempts to consolidate its services have also stumbled. In July, the company merged several ChatGPT modes into a single “super app” and renamed the classic interface, prompting user complaints about cluttered design; the change was quickly reversed. Separate products such as the ChatGPT Atlas browser, the Operator web-agent, and the video-sharing platform Sora were each retired within a year, with key features folded back into the main ChatGPT experience.

Apple’s AI-enhanced Siri, announced as part of Apple Intelligence in 2024, faced repeated postponements attributed to engineering flaws and software bugs. The delays contributed to a $250 million settlement over alleged misrepresentation of AI capabilities in the iPhone 16. Siri’s AI version finally appeared in the iOS 27 beta in July, initially limited to English users with broader rollout planned.

Microsoft introduced the Recall feature for Windows PCs in 2024, promoting it as an AI-powered “photographic memory” that periodically captured screenshots for later search. Immediate privacy concerns emerged, citing the risk of storing passwords, messages, and financial data. After a year-long delay to address security, a researcher demonstrated a tool that could still extract captured information, keeping the controversy alive.

Several hardware and niche software ventures have also folded. The Humane AI Pin, a wearable that raised $230 million, was discontinued in February 2025 after performance problems and a battery-fire warning, with assets sold to HP for $116 million. Rabbit’s R1 companion, despite early sales of 100 000 units, faced criticism for unreliable performance and is now being shifted toward a new “Project Cyberdeck.” Additional closures include the audio app Huxe, model-comparison platform Yupp, role-play service Figgs AI, and Notion Mail, each citing market mismatch or cost pressures.

Sources

  1. The AI graveyard: a running list of projects and startups that didn't make it TechCrunch

More reports

Industry · October 25, 2026 · 2 min

Study Finds Grok Favors Republicans While OpenAI Model Leans Democrat

When forced to name a candidate, SpaceXAI's Grok 4.7 chose Republicans 92% of the time, whereas OpenAI's GPT-6.1 Sol selected Democrats in 97% of cases, according to a Convovo analysis.

Industry · October 25, 2026 · 1 min

BBC’s Blue Planet III Embraces Imperfection to Counter AI-Generated Polish

The new nature documentary deliberately leaves unedited, grainy footage in place, aiming to show a more genuine ocean view amid growing concerns over AI-enhanced visuals.

Industry · October 25, 2026 · 2 min

Pennsylvania law mandates labels on AI-generated advertisements to curb deceptive practices

Governor Josh Shapiro signed the AI Deceptive Advertising Act, obligating all AI-altered ads,image, video, audio or text,to display a matching label, while leaving terms like “significantly modified” and “evidently fictional” vague.