AI Boom Leaves Hundreds of Pre-ChatGPT Startups Stranded as “Fallen Unicorns”

Generative AI has redirected more than $250 billion toward labs like OpenAI and Anthropic—while hundreds of U.S. startups built before ChatGPT’s 2022 debut are stuck with stale billion-dollar valuations, shrinking private-market marks, and few paths to fresh funding or IPOs.

PitchBook data shared exclusively with CNBC show more than 220 “fallen unicorns”—companies that once hit $1 billion valuations but no longer qualify—alongside sharp markdowns for firms that last raised in 2021 or 2022.

The Scale of the Unicorn Hangover

There are 857 U.S. startups valued at $1 billion or more, PitchBook reported. Nearly half have not raised new funding in three years, leaving marks potentially stale.

• Startups that last raised in 2021: down 68% on average (PitchBook estimates)
• Those that last raised in 2022: down 52%

Public software giants such as Salesforce, ServiceNow, and Workday have sold off on AI disruption fears; CNBC describes a quieter but parallel reckoning in private markets.

From Pandemic Boom to ChatGPT Reset

Five years ago, venture capital flooded direct-to-consumer brands, fintech apps, and enterprise SaaS at nosebleed multiples—supported by cheap money and pandemic demand. Even after the Fed began raising rates in 2022, many founders expected to grow into those marks.

Then ChatGPT reframed the landscape.

“The ChatGPT moment was when people said, ‘Holy smokes, the next generation of entrepreneurs, their coding language is spoken English,’” Khosla Ventures partner Samir Kaul (an early OpenAI backer) told CNBC.

“Now you’re seeing 50 engineers do what it would’ve taken 500 engineers to do five years ago,” Kaul said. “We had to completely reshuffle how we valued these companies.”

Who’s on the Fallen-Unicorn List

PitchBook’s fallen-unicorn roster includes consumer and fintech names once ubiquitous in marketing and podcasts:

• Glossier, Brooklinen, Rothy’s, Savage X Fenty, The Farmer’s Dog
• AG1, Betterment, SeatGeek

Enterprise software is the largest category: 75 SaaS firms on the list—double the number of fintech companies—reflecting both 2021 peak multiples and AI’s threat to per-seat workflow tools.

Former DoorDash engineering leader David Zhu, now founder of sales-and-marketing automation startup Reevo, told CNBC his thesis is blunt: workflow-driven enterprise SaaS will be “either disrupted or dead in the next decade.”

Pre-AI companies carry bloated staffing and product debt, Zhu argued: “Unless they make a stark, 180-degree pivot to rebuild the exact same thing from scratch, they’re going to slowly fail.”

Why Funding Dried Up

Mercury CEO Immad Akhund—whose company raised $200 million last month and banks a third of early-stage U.S. venture-backed firms—said pre-AI startups are “definitely in a difficult spot.”

“All the attention’s on AI, so if you’re not an AI-first company, you need really strong numbers to raise,” he told CNBC.

PitchBook analyst Andrew Akers said companies that have not raised since 2021 or 2022 are unlikely to do so again; without venture money or a credible IPO path, many may exit via acquisitions at steep discounts.

“Underneath the surface, I think there are a lot of dominoes to fall,” Akers said.

Discount Acquisitions and a Collapsing Valuation Floor

CNBC cited recent reset deals:

• Stash acquired by Singapore’s Grab in February 2026 at roughly $425 million enterprise value—below ~$660 million lifetime investor capital
• Step acquired by YouTuber MrBeast for an undisclosed sum investors believe was far below ~$500 million prior funding

Restive Ventures’ Ryan Falvey said valuations have compressed about sixfold from 2021 peaks of ~50x forward revenue—a company with the same revenue might be worth ~85% less today.

During the boom, acquirers often paid ~$2 million per engineer as a valuation floor. AI coding tools shattered that assumption by letting smaller teams ship faster—shrinking exit options for older startups.

Post-ChatGPT Winners—and the Investor Question

Falvey said investments made after ChatGPT have been “undoubtedly the best” for his firm; by 2023, post-GPT portfolio companies were already out-earning many pre-GPT bets.

Kaul said he now asks every legacy software pitch: “Why can’t OpenAI, Anthropic or Google do this?” For most, he answers: “They can.”

Survivors may need outcome-based pricing and AI-native infrastructure rather than per-seat SaaS models tied to headcount—models AI agents threaten to compress.

Not every markdown is final: drone maker Skydio disputed PitchBook’s estimate (from $2.5B to $509M) and later announced $110 million from existing investors at a $4.4 billion valuation. AG1 was reported exploring a sale at about $2 billion including debt after CNBC inquiries, per Reuters coverage cited in the piece.

Bottom Line

The AI capital wave is not just lifting new labs—it is repricing an entire generation of venture-backed companies built for a pre-agent world. Fallen unicorns, stale marks, and discount M&A are the private-market mirror of public SaaS fear. For founders and LPs, the lesson is structural: in 2026, “unicorn” status from 2021 is not a floor—it is often a legacy label waiting for a reset.

Frequently Asked Questions