For a generation of startups built before and around the rise of ChatGPT, the lofty valuations they once enjoyed have quickly given way to far more modest expectations.
According to PitchBook, nearly half of 857 U.S. startups once valued at $1 billion or more have not raised fresh funding in three years.
Five years ago, venture capitalists were pouring money into startups selling everything from scheduling software and customizable dog food to lingerie and online ticketing platforms, often assigning billion-dollar valuations before those companies had demonstrated sustained profitability.
Today, more than 220 of those once-celebrated companies are now considered “fallen unicorns.”
In the aftermath of the pandemic, a combination of cheap money and pent-up demand helped fuel a startup boom. PitchBook defines a unicorn as a privately held company valued at $1 billion or more, and 857 U.S. startups reached that threshold.
But by the end of 2025, nearly half of them had gone at least three years without raising new capital.
Startups that last raised money in 2021 were estimated to be worth 68% less on average by the end of last year, while those that last raised in 2022 had seen an estimated 52% decline, according to PitchBook.
The list includes a number of familiar consumer brands, including Glossier, Rothy’s, Brooklinen, Savage X Fenty, The Farmer’s Dog, AG1, Betterment and SeatGeek.
Many of those companies came of age in an environment that rewarded extremely high valuations based on two broad assumptions: interest rates would remain low and startups could always find buyers willing to pay for their technology or engineering talent.
The arrival of ChatGPT in November 2022 changed those assumptions dramatically.
“Now you’re seeing 50 engineers do what it would’ve taken 500 engineers to do five years ago,” said Samir Kaul, a partner at venture firm Khosla Ventures. “We had to completely reshuffle how we valued these companies.”
As the AI boom has directed enormous amounts of capital toward companies such as OpenAI and Anthropic, many older startups have found themselves in a difficult position.
Some are stuck with valuations set during the peak of the venture-capital boom, while their technology and staffing models were built before generative AI dramatically changed the economics of software development.
At the same time, many remain too unprofitable or too slow-growing to attract public-market investors.
“A lot of those companies are pre-AI, not just in their cost structure, but also in their products,” said Mercury CEO Immad Akhund, whose company provides banking services to a large share of early-stage U.S. venture-backed firms.
“They’re definitely in a difficult spot,” Akhund said. “All the attention’s on AI, so if you’re not an AI-first company, you need really strong numbers to raise.”
Enterprise software companies have been among the hardest hit.
Software-as-a-service firms represent the largest category on PitchBook’s fallen-unicorn list, with 75 SaaS companies included—roughly twice the number of fintech companies, the next-largest group.
Many companies founded before the generative AI boom are now burdened by larger staffing models and software architectures designed for a very different technological environment, making a major transformation difficult.
Before the reset, startups could often be acquired by larger companies primarily for their engineers, with acquisition values sometimes estimated at roughly $2 million per developer.
That floor has weakened considerably as AI-assisted coding tools have allowed smaller teams to build increasingly sophisticated products.
As a result, there are fewer obvious exit opportunities for startups whose primary value once rested in technical talent.
Still, the term “fallen unicorn” can be misleading.
PitchBook’s estimates are based partly on employee growth and comparisons with publicly traded companies, rather than actual new financing rounds. Many of the startups identified remain in business, continue to grow and, in some cases, remain profitable.
Some of the companies have also pushed back strongly against PitchBook’s conclusions.
Drone maker Skydio recently announced a $110 million funding round that valued the company at $4.4 billion.
Supplement company AG1 has reportedly explored selling all or part of its business at a valuation of roughly $2 billion.
A spokesperson for The Farmer’s Dog also disputed PitchBook’s estimate, saying the company’s valuation has “only increased since our last funding round in 2022.”
There is also growing concern that today’s AI-native companies may themselves be receiving valuations that resemble the excesses of the 2021 SaaS boom.
With enormous amounts of venture capital flowing into a relatively small group of AI companies, some observers warn that the market could face another painful correction if AI models become increasingly commoditized or if computing costs fail to improve as expected.
The broader lesson is that valuations do not necessarily collapse because companies suddenly become worse businesses.
They can collapse because the benchmark for what is possible changes dramatically.
For startups whose business models were built around assumptions about software development costs, staffing levels and workflows that predate ChatGPT, the challenge may be less about traditional competition and more about adapting to an entirely new cost structure.
