When venture capitalists wrote their largest checks in 2025, they weren’t betting on the next social media platform or delivery app. They were investing in artificial intelligence—and doing so at a scale that has fundamentally reshaped the entire technology investment landscape. According to new analysis from the OECD, AI firms captured 61% of global venture capital investment last year, a figure that has more than doubled since 2022 when AI accounted for just 30% of VC funding. “The concentration of capital in AI represents one of the most significant shifts in venture investing since the dawn of the internet era.” — OECD Directorate for Science, Technology and Innovation The $258.7 Billion Question The numbers are staggering. AI firms raised $258.7 billion out of a total $427.1 billion in global venture capital during 2025. To put that in perspective, AI investment alone last year exceeded the GDP of countries like Portugal, New Zealand, or Vietnam. The infrastructure play has become the dominant investment thesis. Since 2023, AI firms focused on IT infrastructure and hosting have attracted the largest volumes of VC, reaching $109.3 billion in 2025 alone. Cumulative investment in this sector totaled $256.1 billion between 2012 and 2025—meaning nearly half of all infrastructure investment in over a decade came in just the past two years. The geographic concentration tells its own story. United States-based firms attracted about 75% ($194 billion) of global AI VC deal value. The EU27 captured 6% ($15.8 billion), China took 5% ($13.9 billion), and the United Kingdom secured 5% ($13.8 billion). This concentration raises questions about whether other regions can compete in the AI infrastructure arms race. “We’re seeing a winner-take-most dynamic where the companies that can access capital at scale are pulling away from the competition.” — Venture Capital Analyst The Rise of Mega Deals Perhaps the most striking trend is the shift toward enormous funding rounds. AI VC investments are increasingly concentrating in “mega deals” exceeding $100 million. Since 2023, the share of AI VC coming from early-stage investment rounds relative to later funding rounds has been declining. Mega deals now account for about 73% of total AI investment value in 2025. Of these mega deals, transactions above $1 billion represent roughly half of total AI investment value—indicating a market where only the best-capitalized players can compete. The implications for startups are profound. Early-stage AI companies may find it increasingly difficult to raise capital as investors concentrate their firepower on established players with proven traction. This could create a two-tier market: mega-funded incumbents and capital-constrained challengers struggling to reach scale. The infrastructure bottleneck explains much of this concentration. Training large AI models requires enormous computational resources, and the companies that can afford the most GPUs, the largest data centers, and the biggest research teams are pulling ahead. The $256 billion cumulative investment in AI infrastructure reflects a belief that compute capacity will be the defining competitive advantage of the next decade. What Comes Next The OECD analysis comes with a note of caution: venture capital markets are cyclical. While long-term prospects for AI remain strong, the current investment frenzy may not be sustainable indefinitely. The report notes that these findings point to the scale and direction of recent investment trends, but should be interpreted with caution when assessing future developments. For policymakers, the concentration of AI investment in the United States raises strategic questions. As AI becomes increasingly central to economic competitiveness and national security, other regions may need to reconsider their approaches to supporting domestic AI development. For investors, the question is whether the current valuations can be justified by future returns. With half of all AI investment value flowing into billion-dollar-plus deals, the pressure on these companies to deliver massive outcomes has never been higher. One thing is certain: the AI investment landscape of 2025 looks nothing like it did just three years ago. Whether this represents a new normal or a peak before a correction remains to be seen. This article was reported by the ArtificialDaily editorial team. For more information, visit OECD. Related posts: As AI data centers hit power limits, Peak XV backs Indian startup C2i What’s next for Chinese open-source AI Claude Code costs up to $200 a month. Goose does the same thing for fr What’s next for Chinese open-source AI Post navigation Claude Code costs up to $200 a month. Goose does the same thing for fr Railway secures $100 million to challenge AWS with AI-native cloud inf