If you’re building anything that isn’t an AI company, also expect a bumpier road to venture capital. New numbers from PitchBook, referenced by several market watchers, indicate that AI is soaking up most of the new VC dollars — elbowing aside categories that previously dominated just a few years ago.
The imbalance is stark. In total, AI companies have raised an estimated $192.7 billion this year out of about $366.8 billion in global venture investment, according to PitchBook. Last quarter, AI was the recipient of approximately 63% of U.S. VC dollars and slightly more than 53% globally — a phenomenon not seen in modern venture history.
AI Gobbles Up A Record Slice Of Venture Dollars
The bulk of the work is being done by mega-rounds. Anthropic’s $13 billion Series F highlighted how at the frontier model developers can raise like infrastructure providers. Funds of approximately the same scale have coagulated around a few generative AI platforms, GPU cloud companies and model-serving startups with the backing of hyperscalers and sovereign wealth funds.
That capital concentration can be seen even in those that do survive, from fund formation as well. So far this year, there have been just 823 venture funds closed globally, according to PitchBook data — a sharp decrease from more than 4,000 in 2022. Less money and bigger checks for a smaller set of AI names translate to less dry powder for everything else.
Investors paint a picture of a market where AI is not simply a theme but part of the capital stack. Limited partners are directing allocations to managers believed to have special access to AI, and those managers are saving large follow-ons for platform victors. The upshot: a feedback loop that boosts AI’s share.
A Barbell Market That’s Rewarding Giants
Analysts at PitchBook describe today’s environment as barbell-shaped: capital is pooling in and around the largest AI logos and the very biggest companies. Mid-market and emerging managers — usually the lifeblood for non-AI niches — are raising more slowly, writing fewer first checks or not cutting any at all.
Public-market signals amplify the tilt. The Bessemer Cloud Index shows software multiples off their pandemic highs, as AI infrastructure names tied to the GPU supply chain have garnered premiums. When exit comps are friendly to AI, private valuations follow suit and partner time and IC attention shift accordingly.
It’s getting harder for non-AI startups to raise capital
Founders beyond AI haven’t been shut out, but the goalposts have shifted. Deal teams will ask for capital efficiency, gross margins that can withstand a higher-rate world and line of sight to profitability. Sales cycles are being watched, and growth-at-all-costs is out of favor unless there’s an AI advantage that underpins the product.
Investors are still writing term sheets in areas like climate tech, digital health, fintech and supply chain tools — but often for flatter valuations, or stricter milestones or more structure. VCs who were once sponsoring “category creation” are calling out for AI-adjacent roadmaps — an embedded copilot, ML-swung automation, or data network effects — to defend pricing power and raise the ceiling on outcomes.
Further squeezing spending, corporate budgets are tipping toward experiments with AI, rather than the purchase of tools that would have been “nice to have.” In industries with long sales cycles — industrial, health care, public sector — that shift can add quarters to fundraising timelines.
How non-AI founders are responding to tighter funding
Practical teams are repositioning the talking points as specific-to-automation wins, not generic “AI” branding. A warehouse robotics startup, for instance, might measure how much throughput improves with new computer vision upgrades; a payments company might demonstrate to you how they have reduced fraud losses with model tuning. In partner meetings, detail beats slogans.
Capital stacks are changing, too. Founders are coupling smaller equity rounds with venture debt, leveraging non-dilutive sources like SBIR grants, and seeking strategic investments from channel partners. In climate and hard tech, project finance or government-backed programs can be used to fund deployment while saving cap tables for milestone-driven equity.
There’s leverage in the LP community, too. Emerging managers with sector specialization — manufacturing, logistics, healthcare IT — are raising continuation vehicles or annex funds to support winners through a lengthened road to scale. Though more complex to create than a blitzscaling AI round, these structures keep promising non-AI companies on track.
What might rebalance the current flow of venture capital
The pendulum does not always swing.
As model quality standardizes, compression ratios of 20-30x and GPU constraints lighten, more investors will begin to screen for unit economics, customer concentration and defensibility in AI — filters that many non-AI businesses already pass. A reopened IPO window for software and infrastructure beyond AI would also establish new benchmarks and recycle capital to a broader set of managers.
For now, however, the message is unequivocal: AI is the center of gravity in venture. Founders who are out of the blast radius can still raise — more easily than they anticipated but not in a rout (or, at least, only if they fire first) — by delivering crisp efficiency metrics, credible moats and precise ROI stories. The bar is higher — and, as the data suggests, competition for non-AI dollars is stiff.
As a senior analyst at PitchBook said in recent commentary, this market is sharply divided between AI and everything else and between the top funds and the rest. Until that chasm closes, non-AI founders will require conspicuously sharper storytelling and far more creatively structured financing in order to reach yes.
