Historic AI Mega-Round Concentration
A handful of the largest artificial intelligence companies are now soaking up most of the money that investors put into startups worldwide.
What's happening now
In Q1 2026, global venture funding hit a record (about 300 billion dollars by Crunchbase's count, 330.9 billion by KPMG's broader Venture Pulse tally), and AI captured roughly 80% of it. Four deals drove the surge: OpenAI raised a record 122 billion dollars at an 852 billion dollar valuation (Amazon put in 50 billion, Nvidia and SoftBank 30 billion each), followed by Anthropic, xAI, and self-driving firm Waymo, together raising about 188 billion dollars, near 65% of the global total. KPMG counted ten rounds of 2 billion dollars or more worth over 206 billion combined. The concentration kept building into mid-2026: on May 28, Anthropic closed a 65 billion dollar Series H at a 965 billion dollar valuation, briefly passing OpenAI as the most valuable AI startup and signaling a likely IPO. Meanwhile total deal counts fell (North American dollars up 190% year over year but deals down 26%), the classic sign of a barbell market: enormous checks for a few platforms, tighter funding for everyone else.
What it is
A "mega-round" is a single fundraising deal worth billions of dollars, and in early 2026 just four companies, led by OpenAI, raised so much that they alone took close to two-thirds of all global venture investment for the quarter. The result is a lopsided market: a few frontier AI labs get historic sums while thousands of other startups split what is left.
The axis here is whether the historic concentration of venture capital into a handful of frontier AI labs is a rational response to the physics of the problem or a structural entrenchment of incumbents that starves the rest of the ecosystem.
The scale-is-destiny caseFrontier AI is a physical-infrastructure problem before it is a software problem, so concentrating capital in a few labs is the rational, safety-positive, and geopolitically necessary response to a cost curve that genuinely rewards scale.
On this view, training a state-of-the-art model means securing gigawatt-scale energy, custom silicon, and curated data at a scale that cannot be assembled incrementally or split across dozens of undercapitalized labs, and Epoch AI's data showing training compute growing 4 to 5x per year since 2010 makes scale the empirically validated path to capability. The safety argument reinforces it: Anthropic's published core views hold that the most serious alignment problems may only appear in near-human-level systems and that some safety techniques only work on large models, so concentrating capital in a few safety-focused frontier labs produces better outcomes than a fragmented race. The geopolitical case treats mega-rounds as the technology equivalent of defense spending, with U.S. firms controlling roughly 70 percent of global AI compute to China's roughly 10 percent, a lead built by concentrated infrastructure investment that DeepSeek's January 2025 release showed a well-resourced rival can erode quickly.
Sam Altman (OpenAI), Dario Amodei (Anthropic), Masayoshi Son (SoftBank), a16z (Andreessen / Horowitz), and the hyperscaler co-investors (Microsoft, Amazon, Nvidia, SoftBank Vision Fund)
The incumbent-entrenchment caseThe mega-rounds reflect a self-reinforcing flywheel in which preferential cloud credits, exclusive chip allocations, and hyperscaler partnerships let three or four labs raise at valuations no challenger can match, collapsing the diversity of bets on how to build beneficial AI.
The crowding is structural, not anecdotal: in Q1 2026 four frontier companies (OpenAI at 122 billion dollars, Anthropic at 30 billion, xAI at 20 billion, Waymo at 16 billion) captured about 65 percent of all global venture investment while AI overall absorbed roughly 80 percent of the quarter, and mega-rounds took 73 percent of all AI investment value in 2025 with funding increasingly flowing to the same repeat raisers rather than a broadening field. The FTC's Section 6(b) report found that the cloud-AI partnerships, Microsoft with OpenAI and both Amazon and Google with Anthropic, gave hyperscalers equity stakes, revenue-sharing and consultation rights, and access to sensitive model architecture, chip co-design, and customer data unavailable to non-partner developers, creating two classes of AI builder and a regulatory-capture dynamic where compliance costs only incumbents can afford raise the barrier further. The strongest version is not that the labs are malicious (Amodei himself said the concentration happened almost overnight and almost by accident) but that when capital, compute, talent, and regulatory relationships all pool at the same three addresses, there is no independent ecosystem left to correct the error if those labs are wrong.
Former FTC Chair Lina Khan, the FTC Office of Technology (January 2025 Section 6(b) report), the OECD Competition Division (2025), and Anthropic's Dario Amodei (who called the concentration accidental), plus a broad cohort of early-stage VCs and non-AI founders
This is genuinely contested, and honestly so: both camps argue from largely the same facts (compute scaling, the concentration of 2025 to 2026 mega-rounds) and disagree about interpretation rather than evidence. The scale camp has the stronger empirical claim that capability has tracked compute, and serious safety and national-security arguments behind it. The entrenchment camp has the stronger institutional documentation, with FTC staff findings, OECD analysis, and even a frontier-lab CEO conceding the concentration was partly accidental, plus hard concentration numbers showing four deals taking roughly 65 percent of global venture dollars in a single quarter. The disagreement is real because the same data supports both readings: scale may be a law of capability and a mechanism of entrenchment at once.
Shows the concentration extending past Q1: a second frontier lab nearly hit a trillion dollar valuation in May, briefly overtaking OpenAI and pointing toward public-market debuts.
Quantifies the barbell: four companies took near 65% of global venture dollars while overall deal counts fell, the clearest evidence the market is splitting into haves and have-nots.
The largest private funding round in history, at an 852 billion dollar valuation, set the anchor for the entire quarter and showed how much capital a single frontier lab can now command.