North America VC Funding Drops 35% in Q3 as AI Takes Two-Thirds

A smaller quarter, not a weaker market

North American startups raised $92 billion across seed- through growth-stage rounds in the third quarter of 2026, according to Crunchbase data. That is 35% below the prior quarter but 50% above the same period a year earlier 2. The headline decline looks dramatic. Every outlet covering the numbers reaches the same conclusion: the pullback mostly reflects a lack of fresh megarounds for OpenAI and Anthropic, not a broad cooling of investor appetite 123.

That reading is persuasive. When one or two companies can each absorb tens of billions in a single round, quarterly totals become a record of when those rounds close as much as a gauge of venture sentiment. Crunchbase notes that deal volume held fairly steady, with early- and late-stage round counts close to the prior quarter's levels 2. Stable deal counts alongside a falling dollar total point to fewer outsized checks, not fewer investors writing them.

AI's grip tightens

Artificial intelligence again dominated. Roughly two-thirds of North American venture dollars, about $61 billion, went to AI-focused companies 13. ValueAdd VC calls that one of the highest concentrations on record 1. LAVX notes the dollar figure itself is among the largest quarterly AI tallies ever, even though it fell sharply from the two prior quarters 3. Investors kept backing foundation models, infrastructure and applications 3. Large rounds for Databricks, Safe Superintelligence and Crusoe made up a big share of the AI total 2.

The concentration has an obvious downside. ValueAdd VC argues that non-AI startups are fighting over a shrinking slice of a shrinking pie 1. That fits the numbers. When AI takes two-thirds of the money, everything else competes for about $31 billion across the U.S. and Canada.

One caution on the figures: ValueAdd VC's cited totals for earlier quarters don't reconcile cleanly with a 35% sequential drop to $92 billion 1. Crunchbase's own percentages are the safer reference.

Early stage: cooling or holding?

The outlets frame early-stage activity differently. Crunchbase reports a particularly sharp drop in early-stage dealmaking from a multiyear peak in the prior quarter, with later-stage funding also down 2. ValueAdd VC stresses that early-stage funding stayed well above year-ago levels, and reads that as a sign seed and Series A investors haven't retreated despite fatigue at the megaround end 1.

The two accounts aren't really in conflict. Early-stage money can fall hard from an unusual high and still sit comfortably above 2025. The year-over-year comparison seems the better measure of underlying health. By that standard, early-stage investors look committed rather than spooked.

The global picture: fewer dollars, more billion-dollar deals

Worldwide, the pattern holds. Global venture funding fell 25% from Q2's $212 billion but rose 53% from the $104 billion raised in Q3 2025 4. The quarter also set a record for the number of billion-dollar rounds 4. Databricks and Safe Superintelligence each raised $5 billion. Crusoe, Moonshot AI, Mistral AI, Nscale, The Boring Co. and Kling AI each raised more than $3 billion 4.

Global late-stage funding reached $105 billion, down 23% from Q2 but up 73% year over year 4. Even seed is being reshaped by giant checks. Of $13 billion in global seed funding, $2.6 billion came in seed rounds of $100 million or more 4. A nine-figure "seed" round says a lot about how much capital AI founders can now command from day one.

Exits remain the weak link

The funding story is resilient, but the exit story is not. Only 17 venture-backed North American companies went public in Q3, raising under $4 billion combined, with no blockbuster listing 1. Mergers and acquisitions offered more relief. Buyers closed 11 acquisitions above $1 billion, led by Nvidia's $12.93 billion purchase of Hugging Face 3.

Crunchbase's headline points to AI giants eyeing public markets 2. That may be the most important thread. Investors have poured unprecedented sums into a few AI companies at rising valuations. Eventually that capital needs liquidity. Strategic acquisitions by cash-rich incumbents like Nvidia can return some of it, but they can't absorb companies of OpenAI's or Anthropic's scale.

The takeaway

The 35% drop is mostly noise from megaround timing. The more lasting signals are elsewhere: AI's near-total command of venture capital, early-stage investors who remain active, and an IPO market that hasn't reopened for venture-backed companies. If the largest AI companies do move toward listings, they could test whether public investors will validate private-market prices. They could also unlock the liquidity that would let the rest of the ecosystem, AI or not, raise its next round.