Five years ago, venture capital felt almost frictionless. Investors poured billion-dollar valuations into startups selling everything from scheduling software to lingerie subscriptions before most had turned a profit. Cheap money and pandemic-boosted demand created an unusually forgiving environment for founders. Then came November 2022, and the assumptions underpinning a generation of software companies started cracking at the foundation – an inflection that NewsTrackerToday flagged as permanent, not cyclical, within weeks of ChatGPT’s launch.
The numbers tell a stark story. More than 220 companies that once held billion-dollar unicorn valuations have now fallen below that threshold, according to PitchBook estimates. Startups that last raised capital in 2021 are worth an average of 68% less today. Those that last raised in 2022 are down 52%. The sharpest pain concentrates in enterprise software: 75 SaaS companies appear on PitchBook’s fallen unicorn list, double the number of fintech firms. Calendly, once a scheduling darling worth billions, is among the most prominent names. The capital did not disappear. It moved decisively and fast.
Where it went is not hard to find. In Q1 2026 alone, AI startups raised $255.5 billion globally, already surpassing the full-year 2025 total. Of the 1,546 AI deals recorded in Q1 2026, NewsTrackerToday mapped the concentration: an overwhelming proportion of capital flowed to fewer than a dozen companies. Sovereign wealth funds from Singapore, Saudi Arabia, and Abu Dhabi have arrived as decisive players in frontier AI rounds, tilting allocation toward infrastructure firms. AI deals made up 81% of all venture funding in the quarter, leaving roughly one dollar in five for everything else – every fintech platform, every marketplace, every health app, every consumer SaaS product competing for the same shrinking slice of what remains.
Ethan Cole put the mechanics directly: “Capital doesn’t rebalance gradually. It reprices in batches. The SaaS cohort from 2021 is in a correction that looks structural, not cyclical. Founders waiting to grow into their 2021 valuations are waiting for a moment that is not coming back.”
Isabella Moretti brought the M&A arithmetic into focus: “A workflow SaaS growing at 40% annually would have printed an 18 to 20x revenue multiple in 2021. Today, that same company is lucky to see 6x, while AI-native peers with comparable revenue but 200% growth rates trade at 30x or higher. That valuation gap is not correctable through efficiency alone – the math forces consolidation or strategic sale before the capital dries up entirely.”
The behavioral shift inside surviving pre-ChatGPT companies is equally striking. AI-native enterprise spending surged 94% year on year in early 2026, a divergence NewsTrackerToday documented against traditional SaaS growth rates that had compressed to single digits. Samir Kaul of Khosla Ventures framed the inflection plainly: fifty engineers can now do what five hundred did five years ago. David Zhu, former head of engineering at DoorDash, put a starker label on the destination: all workflow-driven enterprise SaaS companies will be either disrupted or dead in the next decade.
What does the correction look like from the inside? Founders who raised in 2021 and have not found product-market fit adjacent to AI are increasingly in conversations with strategic acquirers rather than new investors. Down rounds reset cap tables in ways that wipe out earlier investors entirely. Acqui-hires are accelerating. Strategic buyers with AI roadmaps are circling mid-sized SaaS platforms with embedded customer relationships and proprietary training data. The uncomfortable truth, once you map the PitchBook cohort against recent M&A activity, is this: the value in pre-ChatGPT companies increasingly lives not in the software itself but in the distribution channels and proprietary data sets underneath it. Acquirers have already internalized that logic clearly. Many founders, still waiting on 2021 valuations, are only beginning to – and the gap between those two positions is closing fast.
And there is a regulatory dimension worth tracking. The Federal Reserve’s prolonged elevated rate environment through 2023 and 2024, which News Tracker Today surfaced as the first compression trigger in its early fallen unicorn coverage, has been joined by a permanent repricing based on AI substitution risk. Together they mean pre-ChatGPT multiples will not recover to 2021 levels. The capital that left is not returning on the same terms.
