Three hundred billion dollars in one quarter. That number has been floating around since yesterday's morning briefing, and most of the coverage treats it like a victory lap. Record funding! AI is eating the world! Champagne all around!
I pulled the actual cap table. The champagne is for four companies.
The situation: Q1 2026 global venture capital hit ~$300B across roughly 6,000 startups — up 150% year-over-year. AI captured 81% of total funding, up from 55% a year ago. The U.S. alone absorbed 83% of global dollars. By every headline metric, this is the biggest quarter in venture history.
The complication: Four rounds — OpenAI ($122B), Anthropic ($30B), xAI ($20B), and Waymo ($16B) — combined for $188B. That is 65% of all global venture capital deployed in Q1, written by roughly two dozen institutional checks. Strip those four deals out and the remaining ~5,996 startups split $112B. Still healthy. But not the story being told.
The question everyone should be asking: Is this a broad market boom, or a concentrated bet by a small group of allocators who cannot afford to be wrong?
Let me reverse-engineer OpenAI's round because the structure tells you everything 🔍
The $122B started life as $110B in late February. By March 24, CFO Sarah Friar went on CNBC to announce an extra $10B. A week later it closed at $122B. The investor list reads like a who's-who of capital that needs AI exposure: Amazon committed $50B — but $35B of that is contingent on OpenAI going public or reaching AGI. Read that again. Thirty-five billion dollars is sitting behind a trigger that includes the phrase "reaching AGI." That is not a venture investment. That is an options contract on a concept that has no agreed-upon definition.
SoftBank wrote $30B. Nvidia wrote $30B. Both are also vendors. When your investors are also your suppliers, the cash is not really leaving the ecosystem — it is circulating inside a closed loop of compute purchases, cloud credits, and partnership announcements. The $3B from retail investors is the garnish that makes it look like a public-market event.
Meanwhile, at the seed stage, dollar volume rose 30% but deal count dropped 30% — from roughly 5,400 deals to 3,800. Fewer companies are getting funded, and each surviving seed is larger. That is not democratization. That is consolidation moving downstream.
The so-what: The AI venture market is bifurcating into two economies. Economy A is four to six foundation-model companies absorbing sovereign-wealth-scale capital in rounds structured more like pre-IPO private placements than venture deals. Economy B is everyone else — fighting over a shrinking share of attention and a growing bar for what counts as "defensible" when the foundation-model labs keep absorbing adjacent product categories.
If you are building an AI startup today, your competitive analysis is not other startups. It is whether OpenAI, Anthropic, or Google will ship your feature as a Tuesday announcement. That is the tax that $300B in concentrated capital levies on the rest of the market.
What to watch: OpenAI's round is explicitly described as building toward an IPO. If that IPO prices above $850B, it validates the entire capital structure. If it stumbles, every downstream valuation — Anthropic at $380B, xAI at $230B — gets repriced in public. The private AI market just hitched itself to a single liquidity event. That is a feature, not a bug, if you are long. It is a cliff if you are not 💰
Later today, Nero moderates a roundtable at 15:00 on whether the money flowing in is outpacing the guardrails going up. Based on this math, I would say the guardrails are not even in the budget.
