Introduction
The data point to a market that is becoming larger and more specialized at the same time. The biggest checks are flowing toward the foundational layer—models, compute, hosting and data infrastructure—while thousands of newly funded companies are expanding the opportunity set further downstream. For investors, the question is shifting from whether to gain AI exposure to where in the stack that exposure offers the most attractive risk-return profile.
Private investors are not approaching AI through a single trade. LPs are increasing allocations, venture capital is concentrating around infrastructure and foundational technologies, capital is clustering in a small number of ecosystems, and the physical buildout is creating opportunities for infrastructure and credit strategies alongside traditional equity. This report follows that capital from allocator intent to the assets, companies and financing structures receiving it.
AI Becomes the Default Private-Market Growth Theme
Allocator demand provides the clearest starting point for understanding the private AI cycle. S&P Global’s LP survey shows that roughly three-quarters of respondents intend to commit capital to AI, placing it well ahead of healthcare and longevity, security and defence, digital infrastructure and other private-market themes. This is demand at the top of the capital chain: LPs are actively seeking more exposure before managers decide how to deploy it.
The appetite is broad rather than concentrated in one investor type. Multi-family offices and wealth managers show some of the strongest commitment intentions, while endowments also rank AI highly. The dispersion matters because these investors have different liquidity needs, return targets and time horizons; AI is attracting both long-duration institutional capital and private-wealth channels looking for access to growth that remains outside public markets.
AI also cuts across the categories that sit below it in the survey. Healthcare, defence, infrastructure and financial technology increasingly incorporate AI into their own investment theses. That makes AI less a standalone vertical than a technology layer influencing several private-market strategies—and helps explain why allocator interest remains high even as underwriting at the company level becomes more selective.
What the chart tells us
AI is the clear allocation priority. About 74% of respondents intend to commit capital to AI, versus roughly 50% for healthcare and longevity, the next-highest theme.
Private wealth channels are especially aggressive. Multi-family offices approach 90% commitment intent and wealth managers sit in the mid-80% range.
Endowments also show strong conviction, with commitment intent around 80%, reinforcing the appeal of AI to long-duration institutional capital.
The spread across LP types means fund structure matters. Liquidity, stage exposure and sector specialization will influence which allocator groups are most receptive.
AI Becomes the Default Private-Market Growth Theme
Allocator demand provides the clearest starting point for understanding the private AI cycle. S&P Global’s LP survey shows that roughly three-quarters of respondents intend to commit capital to AI, placing it well ahead of healthcare and longevity, security and defence, digital infrastructure and other private-market themes. This is demand at the top of the capital chain: LPs are actively seeking more exposure before managers decide how to deploy it.
The appetite is broad rather than concentrated in one investor type. Multi-family offices and wealth managers show some of the strongest commitment intentions, while endowments also rank AI highly. The dispersion matters because these investors have different liquidity needs, return targets and time horizons; AI is attracting both long-duration institutional capital and private-wealth channels looking for access to growth that remains outside public markets.
AI also cuts across the categories that sit below it in the survey. Healthcare, defence, infrastructure and financial technology increasingly incorporate AI into their own investment theses. That makes AI less a standalone vertical than a technology layer influencing several private-market strategies—and helps explain why allocator interest remains high even as underwriting at the company level becomes more selective.
What the chart tells us
AI is the clear allocation priority. About 74% of respondents intend to commit capital to AI, versus roughly 50% for healthcare and longevity, the next-highest theme.
Private wealth channels are especially aggressive. Multi-family offices approach 90% commitment intent and wealth managers sit in the mid-80% range.
Endowments also show strong conviction, with commitment intent around 80%, reinforcing the appeal of AI to long-duration institutional capital.
The spread across LP types means fund structure matters. Liquidity, stage exposure and sector specialization will influence which allocator groups are most receptive.
Capital Clusters Where the AI Ecosystem Is Deepest
Private AI investment is also geographically concentrated. California stands apart with roughly $218 billion of investment, while Colorado and New York form a distant second tier. The gap shows that AI capital is clustering in ecosystems where technical talent, venture networks, research institutions, hyperscalers and experienced founders are already dense.
The pattern is self-reinforcing. AI companies depend on scarce engineering talent, access to compute, relationships with cloud and semiconductor providers, and deep financing markets. Regions that already possess those capabilities tend to attract more founders and capital, strengthening their advantage through network effects rather than dispersing investment evenly across the country.
For investors, geography therefore affects both access and pricing. Established hubs offer deeper deal flow and management talent, but competition can raise valuations. Smaller ecosystems may provide differentiated opportunities in areas such as defence, healthcare, fintech, industrial software, energy and autonomous systems—particularly where local specialization creates a defensible cluster.
What the chart tells us
California dominates U.S. private AI investment at roughly $218 billion, far ahead of every other state.
Colorado and New York form the next tier at around $19 billion and $13 billion, respectively; Florida, Massachusetts and Texas each attract several billion dollars.
Most states remain below $1 billion, underscoring how concentrated the investable ecosystem still is.
For investors, the trade-off is access versus crowding: established hubs offer stronger networks, while smaller markets may provide more specialized entry points.
The Foundation Layer Absorbs the Biggest Checks
The sector mix becomes even clearer when investment is broken down by focus area. Capital is spreading across data management, cloud computing, healthcare, cybersecurity, robotics and other applications, but the distribution is highly skewed. AI infrastructure, models, research and governance attract by far the largest amount of private investment.
That category rises to roughly $140 billion in 2025, compared with around $40 billion in 2024. Data management and processing is the next-largest focus area at about $30 billion, while every other category is considerably smaller. The largest capital requirements therefore sit closest to the foundation layer, where investors are financing compute-intensive models, data architecture and the systems required to build and govern them.
Below that core, the opportunity set is more fragmented. Healthcare, pharmaceuticals, cybersecurity, AI agents, autonomous vehicles and robotics all attract capital, but generally through smaller checks and more specialized business models. Private investors can therefore choose between concentrated, capital-intensive exposure at the foundation layer and a broader set of downstream application bets with different sector risks.
What the chart tells us
AI infrastructure, models, research and governance reach roughly $140 billion of investment in 2025, dwarfing every other focus area.
Data management and processing is the clear second tier at about $30 billion, reflecting the importance of organizing and processing data for AI systems.
Capital is still broadening into healthcare, cloud, cybersecurity, AI agents and physical AI, but at much smaller absolute levels.
The market is concentrating by dollar value while broadening by use case: the biggest checks stay near the foundation layer, while downstream opportunities multiply.
AI Infrastructure Is Becoming Its Own Asset Class
The physical buildout behind AI creates a different private-market opportunity from venture investing. Goldman Sachs estimates shown here imply hundreds of billions of dollars of AI-exposed hyperscaler capex across the U.S. and the rest of the world. That spending cascades into data centers, power generation, grid connections, cooling, networking and specialized compute—assets that can be financed through equity, infrastructure vehicles and private credit.
This matters because investors do not need to take early-stage technology risk to gain AI exposure. Data-center developments can support real-estate and construction financing; power-intensive campuses create demand for generation and transmission capital; and compute infrastructure can support asset-backed or contracted financing structures. The return profile can therefore be closer to infrastructure or credit than to software venture capital.
The geographic composition is also notable. U.S. public companies account for about $530 billion of AI-exposed capex, while private U.S. companies contribute roughly $35 billion. Outside the U.S., the private component is larger at about $67 billion. The result is a financing market in which public hyperscalers drive much of the spending, but private capital increasingly participates in the assets required to deliver it.
What the chart tells us
U.S. public companies account for roughly $530 billion of AI-exposed hyperscaler capex, compared with about $234 billion from public companies outside the U.S.
Private-company capex is meaningful: about $35 billion in the U.S. and $67 billion globally excluding the U.S.
The opportunity extends beyond corporate equity into data centers, energy, compute, networking and other infrastructure assets.
That creates a second AI return stream for private markets: cash-flow-oriented financing exposure alongside ownership of technology companies.
Company Formation Expands the Opportunity Set
The final part of the investment picture is the pipeline itself. Newly funded AI companies increase from 482 in 2013 to 3,499 in 2025, showing that private capital is not only concentrating into established winners; it is also financing a much larger universe of new businesses. That broadens the set of potential investments across technologies, vertical markets and enabling tools.
Company formation has not moved in a straight line. Funding activity accelerated through the late 2010s, dipped in 2019 and again in 2022, then recovered before a sharp step-up in 2025. The important point is the scale of the endpoint: the ecosystem now produces several times more newly funded companies each year than it did a decade ago.
For investors, more supply is not automatically better. A larger universe creates more opportunities, but also greater dispersion in quality and more pressure on diligence. As the market matures, manager selection, domain expertise and the ability to distinguish durable technical or commercial advantages from capital-driven momentum become more valuable.
What the chart tells us
The number of newly funded AI companies rises from 482 in 2013 to 3,499 in 2025, a more than sevenfold increase.
The market absorbed periodic pullbacks without breaking the longer-term trend: 2022 fell to 1,603 companies before activity recovered in 2023 and 2024.
2025 is a breakout year, with newly funded companies jumping from 2,050 to 3,499 in a single year.
A larger pipeline expands the opportunity set, but it also raises the premium on selectivity, specialist knowledge and underwriting discipline.
Conclusion: AI Is Becoming a Multi-Asset Private-Market Theme
Taken together, the six charts show that private investors are building AI exposure through several distinct channels rather than one monolithic trade. LP demand is strong, venture capital is concentrating around the foundational stack, U.S. investment is clustered in a handful of ecosystems, and hyperscaler spending is turning data centers, power and compute into investable infrastructure. At the same time, rapid company formation is expanding the downstream opportunity set.
The implication is straightforward: the next phase of AI investing will be defined less by gaining generic exposure and more by choosing the right layer of the stack. Venture, growth equity, infrastructure and private credit can all participate, but they underwrite different risks. The investors best positioned to capture returns will be those that match their capital, expertise and time horizon to the specific part of the AI ecosystem they are financing.
Sources & References
OECD. VC Investments in AI. https://www.oecd.org/en/publications/venture-capital-investments-in-artificial-intelligence-through-2025_a13752f5-en/full-report.html
Stanford. AI Index Report 2026. https://hai.stanford.edu/assets/files/ai_index_report_2026.pdf
S&P Global. LP intent.https://www.spglobal.com/market-intelligence/en/news-insights/research/2026/05/the-most-wanted-exposure-lp-allocation-intent-and-unicorn-ai-deal-concentration-in-private-markets
CB Insights. 1Q25 AI report. https://www.cbinsights.com/research/report/ai-trends-q1-2026/
KPMG. VC Investments. https://kpmg.com/xx/en/media/press-releases/2026/07/vc-investment-already-at-five-year-high-of-billions.html
Premium Perks
Since you are an Wealth Stack Subscriber, you get access to all the full length reports our research team makes every week. Interested in learning all the hard data behind the article? If so, this report is just for you.
Want to check the other reports? Visit our website.
