AI Dominates VC Funding: Startups Raised $22.8 Billion in 9 Days as Series A Momentum Builds

Forty-eight hours into the second half of May, 672 venture funding rounds had closed. By day nine, the total had reached nearly $22.8 billion — a pace that suggests founders and investors have moved past the cautious posture that marked the first quarter of 2026.

What is most striking is not the dollar volume, but the concentration. Nearly 58% of all deals went to AI and machine learning startups. That means while the broader market recovered, capital formation became narrower, more focused, and decidedly optimistic about artificial intelligence.

Where the Capital Is Flowing

The largest single funding round in this period was Anduril's $5 billion Series H—a defense technology company building autonomous systems and autonomous aircraft. Mercury, a fintech platform designed as "the bank for AI startups," raised $200 million to hit a $5.2 billion valuation. n8n, a no-code automation platform, saw its valuation double to $5.2 billion following a strategic investment from SAP.

These are not boutique outcomes. They reflect a broader pattern: capital is flowing to companies building tools, infrastructure, and applications for artificial intelligence. The specificity matters. It is not enough to be a software company in 2026; founders understand they must be an AI software company to attract growth-stage funding at speed.

The median deal size for AI rounds in this period was approximately $77 million—well above the broader market median. This suggests that later-stage rounds dominate the AI segment, while earlier-stage funding (seed, pre-seed) is spread across a wider range of sectors.

Series A and B Momentum Signals Growth-Stage Confidence

One often-overlooked signal in funding data is round sequencing. When Series A and Series B rounds accelerate, it means earlier cohorts of startups are graduating from seed funding and moving into scale-up mode. This requires investor confidence not just in the sector, but in the ability of founders to execute and achieve milestones.

In this nine-day window, Series A and B rounds represented 390 of the 672 total rounds tracked. That is 58% of all activity. Seed and pre-seed rounds, while still substantial (150 rounds), represent only 22% of the total. Earlier-stage investors have long expressed caution; the data suggests growth-stage investors are not waiting for further clarity.

Notably, Series C and later rounds (which typically signal mature-stage scaling) comprised 132 rounds, or roughly 20% of the total. This balanced mix—heavy on Series A/B, lighter on very early and very late rounds—suggests a market that is repricing growth and appetite for risk in the AI infrastructure space.

The Geography and Tempo

Funding rounds did not arrive at a steady pace. May 15 saw 152 rounds close; May 19 brought 166 rounds. These two days alone account for 298 of the 672 total rounds tracked, or 44%. Weekend activity (May 17-18) dropped sharply, a reminder that venture capital—despite its global reach—still operates on a five-day business cycle.

The countries attracting capital were diverse. Signals in this period included funding announcements from Switzerland (Neurosoft, brain-computer interfaces), Ireland (Better Futures, AI documentation), Germany (Tiroler ParaStruct, green construction materials), and China (accelerating robotics funding). The narrative that capital is only flowing to U.S. AI companies does not hold; it is flowing faster there, but the distribution is broad.

What Happens Next

The concentration of capital in AI, the acceleration of Series A and B rounds, and the presence of mega-rounds like Anduril's $5 billion suggest one of two outcomes: either the AI boom is sustaining far longer than earlier skeptics expected, or a significant fraction of this capital will reprrice downward in the coming quarters as the market separates sustainable AI applications from speculative ones.

For now, founders with credible AI products and experienced teams face exceptional access to capital. The data does not lie: $22.8 billion in nine days is a vote of confidence.