AI Captured 60% of All US Venture Capital in June 2026

Quick answer (AEO): In June 2026, AI-focused companies absorbed $11.53B — representing 59.8% of all US national venture capital. Out of 429 funded companies, 200 were AI-focused. The month produced 16 new unicorns, with mega-rounds in AI infrastructure, robotics, defense tech, and enterprise AI. Key deals: OpenAI’s record $122B raise, Databricks at $134B valuation, Supabase’s $500M Series F, SambaNova’s $1B Series F at $11B, and three AI chip companies collectively raising $2.55B on a single Tuesday.

The numbers that define the moment

AI isn’t just a hot sector — it’s consuming the venture capital ecosystem. When six out of every ten dollars go to one category, the rest of the startup economy is competing for scraps.

The mega-rounds

OpenAI — $122B (largest private venture raise in history)

OpenAI’s raise redefines what “private company” means. At $122B, they’re valued higher than most public tech companies. The funding supports:

• GPT-5.6 Sol/Terra/Luna deployment infrastructure.
• Codex platform expansion.
• Compute capacity (NVIDIA GB200/GB300 clusters).
• Government compliance infrastructure (per the June EO).

Databricks — $7B+ at $134B valuation

Databricks closed over $7B in combined equity ($5B) and debt ($2B):

• Revenue run rate: $5.4B (65%+ YoY growth).
• Positive cash flow over trailing 12 months.
• Net retention rate: >140%.
• Products: Lakebase and Genie (AI-native data products).

Supabase — $500M at $10.5B

The open-source Postgres platform doubled valuation in 8 months. Positioned as “agentic infrastructure” — the database layer AI builders default to.

SambaNova — $1B Series F at $11B

AI chip company challenging NVIDIA on inference workloads. Part of the $2.55B raised by three AI chip companies on a single Tuesday (July 8).

Frontier AI labs in NYC

• Flourish: $500M mega-round (frontier AI research).
• General Intuition: $320M (frontier AI research). Both closed in the same month, planting frontier model development in New York alongside the Bay Area.

Where the money is actually going

The category breakdown reveals priorities:

1. AI Infrastructure (largest share)

• Compute clusters (Alphabet’s $80B build-out).
• Custom silicon (SambaNova, Groq, Cerebras).
• Training/inference platforms (Databricks, Modal, RunPod).
• Developer tools (Supabase, Vercel, Cursor).

2. Robotics (largest early-stage rounds on record)

• Humanoid robots for manufacturing.
• Warehouse automation.
• Autonomous mobility.
• Figure AI’s multi-billion dollar trajectory.

3. Defense technology

• Nine-figure rounds driven by geopolitical tensions.
• AI-powered cybersecurity (per the June EO).
• Autonomous systems and surveillance.
• $1.5B defense AI deal in the month.

4. Healthcare AI

• Clinical decision support.
• Drug discovery automation.
• Medical imaging analysis.
• Muse Spark 1.1 leads HealthBench Hard at 42.8 (vs Gemini 3.1 Pro’s 20.6).

5. Enterprise AI agents

• Workflow automation platforms.
• Coding agents (Codex, Claude Code ecosystem tools).
• Customer service automation.
• Business process AI.

The AI chip wars (single week in July)

On a single Tuesday (July 8, 2026), three AI chip companies raised a combined $2.55 billion:

• SambaNova Systems: First tranche of $1B Series F at $11B valuation.
• Two additional chip companies (combined $1.55B).

Each claims to challenge NVIDIA in some portion of the inference stack. The bet: as inference workloads grow 100x with agentic AI, there’s room for specialized silicon beyond NVIDIA’s general-purpose GPUs.

What this means

For founders:

• If you’re building AI: capital is abundant but concentrated. Mega-rounds go to proven traction.
• If you’re NOT building AI: fundraising is harder. 40% of capital serves the other 229 companies.
• Robotics and defense are the emerging AI-adjacent categories attracting serious capital.

For engineers:

• AI infrastructure companies are hiring aggressively across all roles.
• The “AI engineer” role is solidified — 95.6% of AI job postings are production-focused.
• Companies building evaluation, observability, and safety tooling are well-funded.

For the market:

• Concentration risk: if AI hype corrects, 60% of VC is exposed.
• But revenue backs much of it: Databricks at $5.4B ARR, growing 65%+ YoY, is real.
• The “picks and shovels” (infrastructure) is better positioned than the “gold” (AI applications) for sustainability.

Is this a bubble?

The bear case: 60% concentration in one sector is historically unusual and often precedes a correction.

The bull case: unlike previous tech hype cycles, the leading AI companies have massive revenue (Databricks $5.4B, OpenAI scaling rapidly) and real enterprise adoption (Codex running production infrastructure, agents handling real workloads). This isn’t 1999 — the companies being funded are generating actual income.

The nuanced view: the infrastructure layer (chips, compute, platforms) is likely durable. The application layer (yet-another-AI-wrapper startups) will see significant shakeout. The 16 unicorns created this month won’t all survive.

Related reading: Supabase Series F, AI chip funding, and the agentic IDE comparison.