Half of All VC Dollars Went to AI in 2025 — Here’s What You Need to Know
$211 billion flowed into AI startups last year, heavily concentrated in the Bay Area and among a handful of mega-round companies
Investors aren’t showing restraint when it comes to AI. According to new data from Crunchbase and HumanX, AI companies pulled in $211 billion in venture capital last year—an 85 percent jump from 2024. Put differently: One out of every two VC dollars went into AI.
The money isn’t evenly spread. Unsurprisingly, it’s heavily concentrated in the San Francisco Bay Area, capturing 60 percent, or $126 billion, of global AI funding, though it accounts for only 22 percent of total global deals. And within the region, 81 percent of all startup capital was allocated to AI businesses, an 11 percentage-point increase from 2024.
Investing Beyond Foundation Models
The blockbuster rounds from Anthropic, OpenAI, Cohere, Perplexity, and others made 2025 feel like a year defined by foundation model giants. But behind those eye-popping numbers—an 180 percent year-over-year increase to $87 billion, nearly 60 percent of VC investment was deployed into areas deemed where people see real AI value. The data reveals that 19 percent of startup funding went to AI infrastructure, such as data labeling and cloud services. Eleven percent was appropriated to robotics and defense technology. Lastly, 15 percent was allocated to AI-driven software, with health and security being the leading sectors.
Female co-founded AI companies are also seeing notable gains. In North America and Europe, they captured a larger share of VC investment last year—47 percent, or nearly $85 billion, compared with 19 percent, or $19 billion, in 2024.
That said, the study’s authors caution against reading too much into total dollars. A handful of unusually large funding rounds by foundation model labs in 2025 skewed the landscape, they claim. Instead, researchers suggest a more consistent measure is the proportion of total funding rounds, which has remained around 20 percent among female-founded companies over the past eight years.
The Region With the Most AI Investment
AI capital continues to pool in the United States. Crunchbase’s data show that 79 percent of AI funding is invested in U.S.-based firms, a three percent increase from the year prior. By comparison, investment in the rest of the world grew by nearly 67 percent between 2024 and 2025.
But as mentioned earlier, much of the money spent on AI businesses is concentrated in one part of the U.S.: the San Francisco Bay Area. Home to Silicon Valley and innovation, it would be more surprising if investors chose to invest their capital elsewhere in the country. Nevertheless, a closer look at the Bay Area shows just how concentrated AI capital has become: of the $126 billion invested in local startups, $113 billion flowed to only 92 companies that raised rounds of $100 million or more, led by none other than OpenAI, Anthropic, xAI, Scale AI, Anysphere, Thinking Machines Lab, and Safe Superintelligence.
2026 AI Predictions
In short, 2025 proved to be a banner year for AI venture capital, driven by headline-grabbing mega-rounds. But beneath the surface, significant investment continues in areas such as infrastructure and applied sectors. There is indeed scale and emerging breadth in the AI ecosystem.
That said, what can we expect this year? Crunchbase predicts we might see more exits from AI companies, either through public markets or through acquisitions. Tracking approximately 6,600 AI companies that have raised more than $3 million since 2023, the company considers more than 2,300 (over half) as “probable or very likely to be acquired.” It also thinks 443 are “probable or very likely to go public.” Crunchbase didn’t disclose any names, however, nor did it explicitly state that these exits would happen in 2026.
Nevertheless, based on these signals, Crunchbase appears comfortable predicting more M&A and IPO activity in the near future.
The release of Crunchbase and HumanX’s AI Funding Report comes months before HumanX holds its second annual conference in (where else?) San Francisco. The AI gathering features decision makers and industry leaders discussing actionable AI strategies. Organizers boast that more than 130 companies (e.g., Databricks, Cerebras Systems, CoreWeave, Runway, Synthesia, Cohere, Inflection AI, Figma, Chime, and Replit) will be featured, having collectively raised over $72 billion since 2018.
“We see capital flowing to companies solving hard problems with durable value, Stefan Weitz, HumanX’s chief executive and co-founder, says in a release. “HumanX is the living ecosystem where these founders and funders come together to stress-test ideas against real-world results. We aren’t just sharing insights on a page; we’re sparking the ‘collisions’ where data turns into action.”