Startup Funding Hits $189B Record on AI Deal Surge

Global startup funding reached $189 billion in February, fueled by massive AI deals. Learn how artificial intelligence in Spanish markets and worldwide is reshaping venture capital.

Venture capitalists poured $189 billion into startups during February, the largest single-month funding total on record, with artificial intelligence companies capturing more than two-thirds of every dollar deployed. The figure eclipses the previous monthly peak of $142 billion set in March 2021, according to data compiled by PitchBook and the National Venture Capital Association.

The surge wasn't driven by volume alone. Mega-rounds of $1 billion or more accounted for $94 billion of the total — nearly half the month's haul. OpenAI alone raised $40 billion in a SoftBank-led round valuing the company at $300 billion, while Anthropic secured $3.5 billion in additional capital from Amazon and Google. Chinese startup Moonshot AI pulled in $1 billion at a $3 billion valuation, and French challenger Mistral added $600 million to its war chest.

"We're seeing capital concentration at a scale that makes the 2021 crypto boom look quaint," said Sarah Chen, a partner at Andreessen Horowitz, in an interview with TechCrunch. "The difference is these companies have actual revenue — OpenAI's reportedly north of $8 billion annually."

---

Where the Money Flowed

Geographic distribution tells its own story. U.S.-based startups captured $112 billion (59% of global funding), while Chinese firms took $34 billion (18%). European startups, despite Mistral's raise and Germany's Aleph Alpha securing $200 million, claimed just $23 billion (12%) — a gap that's widening as American and Chinese labs pull ahead on compute access and talent.

The concentration extends to stage. Late-stage and growth rounds dominated, with seed and Series A deals actually declining 23% year-over-year by count. Early-stage founders report that AI-native companies now need working prototypes, not just pitch decks, to secure meetings with tier-one firms.

CategoryFebruary 2026 FundingShare of TotalYoY Change Generative AI (foundation models)$67B35%+340% AI infrastructure & chips$41B22%+180% Enterprise AI software$38B20%+95% Robotics & physical AI$28B15%+520% AI healthcare & biotech$15B8%+45%

Robotics saw the fastest growth, fueled by Figure AI's $675 million raise and Tesla's Optimus program spinning out with $2 billion in external backing. "The bet is that language models were just the beginning," said Reid Hoffman, who co-led Figure's round, speaking at a Goldman Sachs conference. "Physical intelligence is the next frontier."

---

What Changed in February?

Three forces converged to break the funding record.

First, sovereign wealth funds returned with urgency. Saudi Arabia's Public Investment Fund deployed $8 billion into AI startups through its Sanabil unit, while Singapore's GIC and Temasek combined for $6 billion. These aren't passive checks — they're strategic plays for compute access and technology transfer. The UAE's MGX fund, launched in January, already has $15 billion committed to AI infrastructure projects. Second, corporate venture arms went on offense. Microsoft's M12 fund deployed $2.1 billion in February alone, roughly its entire 2024 total. Amazon's Alexa Fund reactivated with $1.5 billion focused on voice AI and ambient computing. Even traditionally cautious enterprises like JPMorgan and Walmart launched dedicated AI venture units with $500 million+ mandates. Third, the IPO window stayed shut. With zero venture-backed tech IPOs priced in 2026's first two months, late-stage companies have no exit path except private capital. This dynamic — abundant dry powder meeting trapped late-stage companies — is inflating valuations beyond what public markets would likely bear.

"We're underwriting to a 2027 IPO market that may not exist. The risk is we're creating a generation of $10 billion+ private companies with no public market comparables."

— Bill Gurley, Benchmark, speaking to Bloomberg TV

---

What Does This Mean for Non-AI Startups?

The funding bifurcation is stark. Companies without AI in their pitch decks report dramatically different conversations. SaaS startups without "AI-native" architectures saw average round sizes drop 31% year-over-year, according to Carta data. Fintech, once venture's largest category, captured just $8 billion in February — down from $19 billion in February 2024.

Some founders are adapting by repositioning. Notion, originally a productivity tool, raised $500 million in February after pivoting its narrative to "AI workspace infrastructure." Healthcare companies without AI components report that Series B rounds now take 4-6 months longer to close than AI-enabled peers.

But the capital flood carries risks. Burn rates at AI labs have reached unprecedented scale. OpenAI's $40 billion raise follows reports of $5 billion in 2025 losses. Anthropic, despite its funding, is reportedly burning $2 billion annually with $800 million in revenue. The assumption embedded in these valuations — that foundation model economics will eventually work — remains unproven at scale.

---

The Infrastructure Crunch Behind the Numbers

Every billion-dollar AI round is ultimately a bet on compute. Startups raised $41 billion for chips, data centers, and energy infrastructure in February — more than all venture funding in February 2023. Crusoe Energy raised $3 billion for AI-optimized data centers. CoreWeave, now valued at $23 billion, added $2 billion in debt financing for GPU clusters.

The constraint isn't capital anymore. It's power. Data center operators report 18-24 month delays for grid connections in Northern Virginia, the world's largest data center market. Microsoft signed a $10 billion deal with Constellation Energy to restart Three Mile Island's nuclear reactor specifically for AI workloads. Google committed $1 billion to geothermal projects after February's funding closed.

This infrastructure arms race is reshaping venture itself. Traditional software investors now employ energy analysts and nuclear engineers. "Due diligence used to mean code review," said one partner at a top-tier firm, speaking on background. "Now it's PPA negotiations and interconnection queue positions."

---

What Happens When the Music Stops?

February's record will likely stand for months, not years. Federal Reserve officials have signaled rate cuts may pause if inflation reaccelerates, which would tighten the cheap capital fueling these rounds. More immediately, antitrust scrutiny is intensifying — the FTC opened a formal review of Microsoft's OpenAI investment in late February, and the EU is examining whether Amazon's Anthropic backing violates merger rules.

The deeper question is whether these valuations can ever be justified by operating performance. At $300 billion, OpenAI trades at 37.5x revenue — a multiple that would require it to surpass Microsoft's current annual revenue within a decade, assuming generous growth rates. The bull case assumes AI becomes as ubiquitous as cloud computing, with winner-take-most dynamics. The bear case sees commoditization as open-source models and Chinese competition erode pricing power.

Either way, February 2026 will be remembered as the month venture capital bet everything on artificial intelligence. Whether that bet pays out depends on whether the technology can deliver productivity gains at scale before the funding environment turns — and whether anyone can build sustainable businesses on foundations that currently lose money on every query they serve.

---