Global venture funding reached a record $510 billion in the first half of 2026, more than the entire 2025 total, as artificial intelligence swallowed a larger share of the market than ever before. According to Crunchbase data, the surge was powered overwhelmingly by AI — and, within AI, by two companies whose fundraising now distorts the entire venture landscape. Strip them out, and the picture looks far more ordinary.
A Record Half, Powered by AI
The numbers are staggering by any historical measure. The first quarter delivered roughly $305 billion and the second added about $205 billion, making Q2 the second-largest quarter on record. The half-year total surpassed the roughly $440 billion invested across all of 2025.
AI sat at the center of it. AI-focused companies captured more than 70 percent of global startup capital in the second quarter, up from around 50 percent a year earlier. The technology is no longer one sector among many competing for venture dollars — it is the dominant destination for them.
Two Labs, Half the Market
The concentration at the top is extraordinary. OpenAI and Anthropic alone accounted for about $217 billion, or roughly 43 percent of every venture dollar deployed worldwide during the half. Anthropic's second-quarter raise reportedly pushed it past SpaceX to the top of Crunchbase's private-company rankings, and reporting placed its valuation near the very top of the global list, just ahead of OpenAI.
That concentration is the key to reading the data honestly. Once the two frontier labs are excluded, the market looks much more like 2024 and 2025. Crunchbase noted that outside a handful of mega-rounds, activity tracked near recent-year norms — a sign that the headline record is being driven by a small number of enormous deals rather than a broad-based boom across every startup stage.
Enterprise Demand Underneath the Hype
Beneath the mega-rounds, real enterprise adoption is helping justify the flood of capital. Enterprise generative-AI spending reportedly jumped from around $11.5 billion in 2024 to $37 billion in 2025, and longer-range forecasts project the enterprise AI market expanding at a compound annual growth rate near 40 percent over the coming decade.
Notable recent deals reflect that demand for the infrastructure and tooling layer:
- A $1.5 billion financing for an enterprise AI platform focused on model serving.
- A large Series F for an inference-infrastructure company at a multibillion-dollar valuation.
- A $500 million Series D for a voice-AI company, tripling its valuation in under a year.
The through-line is that money is flowing not only to the frontier labs but to the companies that help enterprises actually deploy and run models in production. That "picks and shovels" pattern — funding the serving, tooling and voice layers rather than only the model builders — suggests investors expect the next wave of returns to come from making AI usable at scale, not just from training ever-larger systems.
Discipline Returns to the Deck
Even amid record numbers, investor behavior is tightening. Analysts describe a market that has stopped rewarding AI spending as a signal of ambition and started grading it on attribution — which specific revenue or cost line a given investment moves, and by when. That discipline flows downhill from the public markets, where the Street is asking hyperscalers the same question.
There are warning signs too. Big cloud providers are reportedly planning to spend up to $900 billion on AI infrastructure in 2026, far ahead of current revenue. When capital expenditure runs that far in front of income, history suggests a painful sorting between companies that turn models into profit and those that do not.
Why It Matters
The H1 2026 figures tell a two-sided story about the AI economy.
- The boom is real but narrow. Record totals mask a market that, outside a few giants, resembles prior years.
- Attribution is the new test. Boards and investors increasingly demand proof that AI spending moves a specific business metric.
- Infrastructure over hype. Durable capital is favoring revenue, defensible data and genuine enterprise adoption.
For founders, the easy money for any team with "AI" on a slide is fading. For enterprises, the pressure to show measurable returns — not just experimentation — is rising in lockstep with spending. The likely outcome of this cycle is not a collapse of AI but a sorting: a widening gap between the companies that convert models into profit and those that merely consume the capital chasing them.
