Late-stage venture capital seemed like a dark corridor that nobody wanted to enter for three years. Their deployment timelines were subtly extended by funds raised during the 2021 frenzy. By 2023, founders who had anticipated ringing the bell at the New York Stock Exchange were instead facing fourth and fifth rounds of layoffs, sometimes from offices they had already partially filled. The term “zombie company cohort” that frequently appeared in NVCA reports was remarkably direct for a sector that favors more delicate language. Then, almost without warning, the money returned.
This year’s first quarter figures are the kind that force analysts to double-check their spreadsheets. In a single quarter, global venture investment reached approximately $300 billion, more than in any previous quarter. Of that, $246.6 billion came from late-stage rounds, a 205% increase from the previous year. However, the most telling statistic is hidden a few levels down: only 158 transactions totaling $100 million or more received $235 billion. Nearly 80% of the capital was absorbed in less than 3% of all venture transactions. Yes, the drought was over. However, it ended in a very specific manner.
It’s almost entirely AI. The largest private funding event in history, OpenAI raised $122 billion in a single round, valuing the company at about $852 billion and preparing for an IPO that some bankers believe could clear a trillion. In April, Anthropic took $15 billion. Project Prometheus, Jeff Bezos’s new AI manufacturing venture, raised $10 billion, and xAI began the year with a $20 billion Series E. Companies like Vast Data, Ineffable Intelligence, and Shield AI, which would have made headlines in 2022, are closing rounds that the majority of readers have never heard of.
The rest of the venture economy, meanwhile, appears to be largely unchanged from the lean years. In fact, U.S. venture firms’ fundraising fell 35% to roughly $66 billion in 2025—the lowest amount in at least six years. Due to losses from 2021-vintage funds that have not yielded any returns, limited partners remain cautious. A few sizable multistage firms, including Andreessen Horowitz, Sequoia, Insight, and General Catalyst, are mercilessly consolidating the new money in their hands. These firms have the balance sheets to continue writing $500 million checks to businesses that are still regarded as “private.”
Alongside this, a more subdued change is taking place. In 2025, the initial public offering (IPO) window opened; Circle’s June 5 debut surged over 500% from its offering price, Figma filed, Chime listed, and Klarna finally made its public debut. The first tangible indication that exits weren’t irreversibly damaged came when Jeremy Allaire rang the bell at the NYSE in the early summer and Circle’s market capitalization increased to $42 billion. Late-stage investors are unable to return capital in the absence of exits. The LP flywheel cannot spin in the absence of returned capital. It explains why one successful IPO month can change Sand Hill Road’s entire atmosphere. It’s a small chain of cause and effect.
This recovery isn’t the catch—and there is always a catch. Without an AI component, it is not a recovery for fintech founders. Outside of specific imaging and drug-discovery niches, healthtech isn’t making a comeback. In non-AI categories, Series B and Series C rounds are still competitive, with investors requiring actual profitability rather than projected profitability before issuing growth checks. I’ve heard informally from partners that the post-ZIRP world is still ongoing. The underlying caution has just been obscured by a massive AI capital wave.
Whether this continues is still up in the air. One indicator will be SpaceX’s filing, which is anticipated to gauge interest in trillion-dollar listings. So will the M&A pipeline, which in Q1 resulted in Savvy Games’ $6 billion acquisition of Moonton and Capital One’s $5.15 billion deal for Brex. In the strictest sense, the drought is over. The more difficult question, which is unlikely to be resolved until the IPO class of late 2026 actually begins trading, is whether what has replaced it qualifies as a healthy market or is just a single-sector boom disguised as one.
