AI Eats Two-Thirds of All US Venture Capital

US startups raised a commanding $19.44 billion across 492 deals in July 2026, but the top-line strength masks a profound concentration of capital. Artificial intelligence companies swallowed a staggering $13.08 billion, representing 67 percent of all venture dollars deployed during the month. This massive haul was concentrated in just 196 companies, leaving the rest of the ecosystem to fight over the remaining third of the pool. The high capital-to-deal ratio reveals a market dominated by massive, late-stage rounds rather than broad, early-stage distribution.

Photo by RDNE Stock project via Pexels

This funding landscape highlights a major shift in investor risk tolerance, where safety is found in scale. Geographically, California secured its dominance by capturing $10.49 billion of the monthly total, while traditional hubs like New York saw dramatic declines. New York City startups raised $1.81 billion, marking a steep 54 percent drop from June's record high. Investors are no longer spreading small bets across a wide net of experimental software startups, choosing instead to concentrate their capital into capital-intensive infrastructure.

The data from July indicates that the median deal size has settled at $6 million, but the average deal size in major hubs tells a very different story. In New York, the average deal size actually jumped 40 percent year over year to $34.2 million, even as overall transaction volume slumped by 38 percent. This divergence shows that while fewer companies are getting funded, those that do are receiving unprecedented war chests. Venture capital has effectively transformed from an index of tech innovation into a highly concentrated funding mechanism for machine learning hardware and computing power.

For early-stage founders outside the machine learning pipeline, this funding concentration presents a harsh reality. The capital abundance of the late-stage market is not trickling down to seed and Series A companies in non-AI sectors. Founders must build toward profitability much faster as the bridge to late-stage venture capital narrows. Meanwhile, the pressure on mega-round recipients to deliver outsized returns is mounting, setting up a high-stakes bottleneck for the venture asset class.

Over the next twelve months, this capital concentration will likely force a consolidation wave among mid-tier software-as-a-service startups. As early-stage dry powder remains underutilized, look for corporate development teams to acquire struggling startups for talent rather than technology. The ultimate success of the July cohort will depend on whether these heavily funded AI infrastructure plays can generate enterprise revenue to justify their valuations. If those revenues fail to materialize by mid-2027, the venture market will face a severe correction.