OpenAI and Anthropic soak up half of record $407B AI funding in H1 2026

Venture funding for artificial intelligence startups reached an unprecedented $407 billion in the first half of 2026, a figure that eclipses the $264 billion invested across the entire sector in all of 2025, according to a new report from PitchBook. But beneath the headline number lies a striking imbalance: OpenAI and Anthropic alone absorbed roughly $217 billion—more than half of the global total. The two frontier AI companies have become gravitational centers of capital, pulling in mega-rounds that are reshaping the venture landscape and concentrating power in the hands of a very small number of players.

The concentration is not merely a statistical curiosity. It signals a structural shift in how the world’s largest investors are positioning themselves for an AI-driven economy. When two private companies can raise more money in six months than the entire AI sector did in a record-breaking prior year, it raises fundamental questions about market balance, the viability of smaller startups, and what happens when these giants eventually go public. For executives, founders, and investors globally, the data underscores a stark reality: the AI boom is increasingly a two-horse race, and the ripple effects are being felt far beyond Silicon Valley.

The anatomy of two mega-rounds

OpenAI’s $122 billion round in March 2026 set the tone. Led by Amazon, Nvidia, and SoftBank, the financing was the largest private technology investment ever recorded. Two months later, Anthropic raised $65 billion in a round backed by Altimeter Capital, Dragoneer, Greenoaks, Sequoia, and BlackRock, with strategic participation from Samsung and SK Hynix. The round briefly made Anthropic the world’s most valuable private company at a $965 billion post-money valuation.

These are not typical venture deals. They are capital infusions of a scale once reserved for sovereign wealth funds or entire national industrial policies. Dimitri Zabelin, senior analyst at PitchBook, described the platforms as “structural, not cyclical,” reflecting deep investor conviction that OpenAI and Anthropic are foundational infrastructure for the next decade of computing. The language matters: investors are not betting on a product cycle or a quarterly growth curve. They are betting that these companies will become the operating systems of the AI era, with pricing power and ecosystem control that justify valuations once considered absurd.

The investor rosters tell their own story. Amazon, Nvidia, and SoftBank are not passive financial backers; they are strategic partners with deep commercial relationships. Amazon provides cloud infrastructure through AWS, Nvidia supplies the chips that power model training, and SoftBank has positioned itself as a global AI conglomerate. Anthropic’s backers include semiconductor giants Samsung and SK Hynix, signaling that memory and storage capacity are now as strategically important as compute. These alignments suggest that the mega-rounds are as much about securing supply chains and distribution channels as they are about financial returns.

A market of fewer, larger bets

While total AI funding soared, the number of deals collapsed. PitchBook counted only 3,500 AI deals in the first half of 2026, down from 8,290 in all of 2025. The math is stark: more money is chasing far fewer companies. The average deal size has ballooned, but the pool of funded startups has shrunk dramatically. This is the defining characteristic of the current cycle—a flight to perceived safety at the top of the market, while early-stage and mid-sized AI companies struggle to attract attention.

The concentration is even more extreme when viewed quarterly. In the first quarter of 2026, 67.3% of all AI venture funding went to just three companies: OpenAI, Anthropic, and xAI. The remaining thousands of AI startups globally competed for less than a third of the capital. This dynamic has profound implications for innovation. Frontier AI research is extraordinarily capital-intensive, requiring billions of dollars for compute, data, and talent. But the downstream ecosystem—application developers, vertical AI solutions, tooling providers—depends on a broader distribution of capital to thrive. If that ecosystem is starved, the platforms themselves may eventually suffer from a lack of viable commercial applications built on top of their models.

Some analysts have begun to warn of a market imbalance. The fear is not that OpenAI and Anthropic will fail—they are exceptionally well-capitalized—but that their dominance could crowd out other sectors entirely. Venture capital is a finite resource, and when half of AI funding goes to two companies, other categories—climate tech, biotech, fintech, enterprise software—inevitably feel the squeeze. The PitchBook data suggests that the AI boom is not lifting all boats; it is lifting a few supertankers while leaving the rest of the harbor dry.

The IPO question looms

The concentration of private capital sets up an even more consequential test: the public markets. Both OpenAI and Anthropic are expected to pursue initial public offerings, and their valuations will set benchmarks for the entire AI sector. Anthropic’s IPO valuation, according to reports, hinges on revenue forecasts of $190 to $200 billion by 2028—a projection that would make it one of the most valuable companies in the world shortly after listing. OpenAI’s path is more complicated: its sales growth has reportedly slowed in recent months, unnerving some investors ahead of a planned IPO.

The stakes are enormous. If these companies debut at or above their private valuations, it will validate the mega-round thesis and likely trigger another wave of capital inflows into AI. If they stumble, the correction could be severe, not just for the companies themselves but for the entire venture ecosystem that has piled into the sector. Public market investors are less forgiving than private backers; they will scrutinize revenue quality, margins, and competitive moats with a rigor that mega-round participants have so far been willing to suspend.

For international professionals, the implications extend beyond investment portfolios. The dominance of OpenAI and Anthropic is shaping talent flows, technology standards, and regulatory debates worldwide. Companies in Europe, Asia, and Latin America are increasingly forced to choose between building on these platforms or competing against them—a choice that often determines access to capital, compute, and customers. The concentration of AI power in two American companies also raises geopolitical questions, as governments grapple with how to ensure sovereignty and resilience in a technology that is becoming as foundational as electricity.

What comes next

The PitchBook report captures a moment of extraordinary conviction—and extraordinary risk. The $407 billion invested in AI during the first half of 2026 is a bet on a future that is still being written. OpenAI and Anthropic have the capital, talent, and strategic partnerships to define that future, but they also carry the weight of expectations that would have been unimaginable just three years ago. The coming IPOs will be the first real test of whether the private market’s enthusiasm can survive contact with public market discipline.

For the broader ecosystem, the challenge is to find a path that does not depend on the fortunes of two companies. The AI revolution will ultimately be judged not by the size of the largest rounds, but by the breadth of the economic transformation it enables. That transformation requires a healthy pipeline of startups, a diverse set of applications, and a competitive landscape that rewards innovation rather than incumbency. The first half of 2026 showed that the world’s investors are willing to bet unprecedented sums on AI. The second half will begin to reveal whether those bets are placed wisely.