Global investment in artificial intelligence reached a historic milestone in 2025, with funding nearly doubling year over year to $211 billion, according to a new joint report released by HumanX and Crunchbase. The surge means AI alone now captures about 50% of global venture capital, marking a decisive shift in investor priorities across the tech sector.

The HumanX + Crunchbase AI Funding Report 2025 highlights a market that is no longer chasing experimentation alone. Instead, capital is increasingly concentrated in regions, sectors, and companies demonstrating durable value, operational maturity, and defensible AI capabilities.

The report shows that investor attention has narrowed sharply toward companies operating at scale. Of the $211 billion raised, $163 billion came from megadeals, underscoring a preference for late-stage bets where infrastructure, enterprise adoption, and execution speed are already proven.

Funding to foundation model companies surged 180% year over year, reaching $87 billion, led by firms such as OpenAI and Anthropic. However, the majority of capital flowed beyond core models, with 59% of investment directed toward infrastructure, semiconductors, robotics, security, healthcare, defense, and AI-driven enterprise software, reflecting a shift toward areas where AI is being operationalized and tied more directly to measurable business outcomes.

The report also highlights a notable shift in founder demographics. In North America and Europe, companies with at least one female founder captured 47% of AI funding in 2025, according to Crunchbase data. While AI funding remains concentrated among large investment rounds and established AI hubs, the data suggests broader participation among founding teams as artificial intelligence moves deeper into enterprise deployments, infrastructure-led growth, and production-scale AI systems.

Geographically, the data reveals an unprecedented concentration of AI investment. In 2025, the San Francisco Bay Area accounted for $126 billion in AI funding, representing roughly 60% of all global AI investment. Within the region, AI accounted for 81% of total startup capital, indicating how heavily investor funding in the Bay Area is now weighted toward AI-driven companies.

Despite representing just 22% of global AI funding deals, the Bay Area continues to lead in dollar volume — a sign that fewer, significantly larger investment rounds are driving its outsized share of capital. This reflects investor preference for ecosystems where talent, capital, and enterprise buyers intersect.

Leveraging Crunchbase’s predictive intelligence, the report points to renewed momentum for AI exits in 2026. Of the 138 private companies appearing at HumanX this year, 27 are rated as “probable or very likely” to go public, while another 30 are positioned for acquisition, according to Crunchbase’s forward-looking analysis.

The data suggests that while broad exit activity has slowed since 2022, a subset of AI companies is beginning to separate itself — particularly those demonstrating durable revenue models, deeper customer adoption, and clearer paths to scale. This pattern reflects growing expectations that AI companies demonstrate the same fundamentals required of mature technology businesses, including sustained revenue growth, customer retention, and disciplined cost structures.

“Every AI cycle brings bubble predictions, but the data reveals a much more nuanced story,” said Stefan Weitz, Co-founder and CEO of HumanX. “We see capital flowing to companies solving hard problems with durable value. 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.”

From Crunchbase’s perspective, the emerging exit signals reflect a more disciplined investment environment rather than a slowdown in innovation.

“The AI winners of today are not necessarily the winners of tomorrow,” said Jager McConnell, CEO of Crunchbase. “Crunchbase’s predictive intelligence uses billions of live signals to forecast private market moves, and we foresee that significant AI IPOs are likely in 2026. The data shows that while investors are making more cautious bets, no longer throwing money at anything labelled AI, many of the companies attending HumanX have potential to raise significant rounds this year. We’re only in the first inning of the AI game, and HumanX will be a significant play for many companies.”

These indicators suggest that the next phase of AI exits will be shaped less by hype cycles and more by companies that can demonstrate sustained performance under tighter scrutiny.

While the AI Funding Report outlines where capital is concentrating, HumanX provides a real-world lens into how those dynamics play out in practice. The more than 130 companies taking the stage at HumanX in San Francisco have collectively raised over $72 billion since 2018, according to Crunchbase data — a reflection of how closely the event mirrors the broader contours of the AI market.

This year’s presenters span multiple layers of the AI ecosystem, from industry leaders that have raised significant capital to AI-native innovators and established technology platforms integrating AI at scale.

Companies such as Databricks, Cerebras Systems, and CoreWeave stand out as industry leaders, reflecting continued investor focus on data infrastructure, compute, and performance at scale. At the same time, AI innovators including Runway, Synthesia, Cohere, and Inflection AI highlight where new AI-first applications are gaining traction across media, enterprise software, and conversational systems.

Beyond AI-native companies, market movers such as Figma, Chime, and Replit illustrate how AI is increasingly embedded into broader enterprise and consumer platforms rather than operating as a standalone category.

This range of companies reflects a market that has moved beyond experimentation, as AI becomes embedded in production systems, workflows, and core business operations.

Crunchbase’s forward-looking analysis is not new — and its track record adds context to the 2026 exit signals identified in this year’s report. At HumanX last year, Crunchbase leadership used its predictive intelligence to flag a likely acquisition involving Superhuman, months before any public announcement. In July 2025, that prediction materialized when Grammarly acquired Superhuman.

That outcome was not isolated. Other companies presenting at HumanX have since experienced major financial events aligned with Crunchbase’s earlier forecasts, including the acquisitions of Moonhub, Hour One, and Tome.

Crunchbase’s predictive intelligence is built by analyzing billions of live market signals, drawing from public data, government filings, direct input from investors and employees, thousands of data partners, expert human curation, and aggregated usage data from more than 80 million active Crunchbase users. Since launching predictive intelligence in early 2025, more than 7,000 Crunchbase predictions have been confirmed by real-world events, spanning funding rounds, acquisitions, and IPOs.

In this context, the 2026 exit outlook in the HumanX report is grounded less in speculation and more in observable patterns — drawing on historical outcomes and real-time market signals to identify where momentum is building.

In response to the HumanX + Crunchbase 2025 AI Funding Report, AiNews.com posed a series of questions to HumanX and Crunchbase leadership to better understand how the data translates into real-world investment decisions and exit expectations. Here are their responses.

1. Where is AI capital becoming most concentrated versus where investors are growing more cautious or selective?

“The Bay is running away with the game. Crunchbase's data shows roughly 60% of global AI funding landing in San Francisco in 2025. That's gravitational pull.

Investors are being ruthless about where they place bets. Capital is flowing to places where talent, infrastructure, and speed of execution stack on top of each other. Everywhere else? Fighting for the scraps or the company needs a very specific thesis to justify the check.

The concentration isn't random. San Francisco has the density of researchers, engineers, and enterprise buyers in the same zip code. That flywheel is super hard to replicate, and the data shows investors frankly aren't really trying to.”

2. Is the funding flowing toward genuine AI capability and defensibility or primarily toward scaling existing platforms?

“The wrappers-on-GPT era is over. Crunchbase's data makes it pretty clear that capital is moving toward companies with something proprietary: unique data, real (and safe) infrastructure, actual integration into workflows that matter.

Investors got burned chasing ‘AI-enabled’ pitches that were really just API calls with a logo. The question I ask in venture pitches is: What do you have that OpenAI can’t replicate in six months? What's the moat when the foundation models get even cheaper and faster?

The companies getting funded today are building things that are hard — robotics, security, healthcare, vertical-specific systems where the AI has to actually work in the real world, not just demo well. That's a good correction in my opinion.”

3. What does “exit-ready” actually mean in the current AI market and which signals matter most?

“Exit-ready in AI right now means you've graduated from ‘impressive demo’ to ‘business that is actually a business.’ So things like revenue growth, customer NRR, and falling unit economics.

Crunchbase's predictive data is useful here because it cuts through cherry-picked case studies. The companies tracking toward real exits aren't necessarily the ones with the most hype.

The market is finally asking AI companies the same boring questions it asks everyone else: Do customers pay you? Do they stay? Can you do this profitably? If you can't answer yes to all three, you’re just exit-hopeful.”

1. Where AI capital is becoming most concentrated versus where investors are growing more cautious or selective?

“Funding to the AI sector grew 85% year over year to more than $211 billion, according to Crunchbase data. That represents roughly 50% of total global venture funding last year.”

2. Is the funding flowing toward genuine AI capability and defensibility or primarily toward scaling existing platforms?

“Sectors impacted by AI that showed significant growth year over year and AI concentration include robotics, code automation, enterprise software, defense, legaltech, and cloud data services.

Investors are looking to invest in companies that will define the next era. Our data shows large funding rounds going to a mix of native AI companies growing quickly, as well as established companies realizing new opportunities with AI.”

3. What does “exit-ready” actually mean in the current AI market and which signals matter most?

“Since the market correction at the end of 2021, ‘IPO ready’ means significant revenue (north of $100M to $200M ARR) as well as revenue growth — more than 30% year over year. For M&A, AI talent, product adoption, and differentiation come into play.”

AI funding data is often treated as an investor-only concern, but the patterns revealed in this report have far broader implications. Where capital concentrates determines which technologies get built, which problems are prioritized, and which regions and industries gain influence over how AI is deployed in the real world.

The 2025 funding shift shows that AI has moved out of its experimentation phase. Investors are no longer funding ideas simply because they use AI — they are funding companies that can prove real-world adoption, operational resilience, and long-term value. This affects enterprises deciding which vendors to trust, founders determining what types of products are viable, and policymakers watching where economic power is consolidating.

HumanX and Crunchbase published this report to move the conversation beyond hype cycles and headline-driven narratives. By combining funding data with predictive intelligence and on-the-ground signals from companies presenting at HumanX, the report offers a clearer picture of which AI businesses are likely to endure — and why.

For readers, the takeaway is not that AI’s future is out of reach — but that it is becoming more defined. Capital is concentrating where execution is fastest and ecosystems are densest, yet the report also shows that companies earning sustained investment are those building defensible systems, solving real problems, and proving long-term value. For founders, enterprises, and technologists outside traditional hubs, these signals clarify what investors now reward — and where differentiated approaches, deep domain expertise, and disciplined execution can still break through.

In today’s AI market, funding is no longer about belief — it’s about proof.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.