AI Kills the Unicorn: How ChatGPT Wiped Out Hundreds of Billion-Dollar Startups

The post-pandemic venture boom minted unicorns at a furious pace, but generative AI has quietly turned that era's crowning achievement — the billion-dollar valuation — into a millstone. Of the 857 U.S. startups currently holding unicorn status, nearly half haven't raised fresh funding in three years, leaving their valuations effectively fictional. PitchBook estimates that startups last funded in 2021 are now worth 68% less on average, while 2022-vintage companies have shed roughly 52% of their value. More than 220 have officially crossed into 'fallen unicorn' territory.

The fallen include some of the most heavily advertised brands of the early 2020s: Rihanna's Savage X Fenty, beauty brand Glossier, pet food company The Farmer's Dog, shoe brand Rothy's, supplement maker AG1, robo-advisor Betterment, and ticketing platform SeatGeek. These companies were built on two assumptions that no longer hold: that interest rates would stay near zero indefinitely, and that larger tech companies would always pay roughly $2 million per engineer to acquire talent. AI coding tools have demolished that talent-acquisition floor — smaller teams can now build what once required hundreds of engineers, making acqui-hires far less attractive.

Enterprise SaaS companies are the single hardest-hit category, with 75 software-as-a-service firms appearing on PitchBook's fallen unicorn list — double the number of fintech casualties. The logic is brutal: SaaS business models are built around charging per user seat, embedding software into employee workflows. Autonomous AI agents threaten to make those workflows — and the headcount driving subscription revenue — largely obsolete. David Zhu, a former DoorDash head of engineering who led 200-plus engineers, now argues that all workflow-driven enterprise SaaS companies will be either disrupted or dead within a decade. He founded Reevo, an AI platform replacing corporate sales and marketing teams entirely.

The valuation compression is staggering in magnitude. Revenue multiples have collapsed roughly sixfold from the 2021 peak of 50x forward revenues, meaning a company with identical revenue is worth approximately 85% less today than five years ago. Recent acquisitions illustrate the carnage: savings app Stash sold to Singapore's Grab for $425 million against roughly $660 million in total investor capital. Fintech Step was acquired by YouTube star MrBeast at a price investors widely believe came in well below the $500 million the startup had raised. Without a viable IPO path or fresh venture funding, distressed fire-sale acquisitions are likely the best outcome most fallen unicorns can hope for.

The deeper structural shift may be permanent. Post-ChatGPT startups are already generating more revenue than comparably aged pre-AI companies, according to fintech-focused Restive Ventures. Khosla Ventures partner Samir Kaul — whose firm was an early OpenAI backer — now applies a single litmus test to every incumbent software company seeking investment: could OpenAI, Anthropic, or Google simply build this themselves? For most, the honest answer is yes. The venture industry is recalibrating around outcome-based pricing, AI-native infrastructure, and radically leaner teams — leaving the unicorn class of 2020-2022 with little room to maneuver.