How will the AI infrastructure boom be financed?

Key takeaways

• The historic scale of AI infrastructure investment is driving capital demand across different asset categories, and anticipated mega IPOs represent the tip of the iceberg supporting AI growth investment.
• The resulting investment opportunity incorporates high growth technology companies and the related ecosystems, and builders and operators of the AI infrastructure, creating investment opportunities across sectors and asset classes.
• The historically unprecedented AI infrastructure buildout will need financing from any and all sources, accessing funding from public and private markets across credit, equity, and structured financing, and the hyperscalers also deploying their corporate cash flow.
• A multi-manager approach combines specialized expertise across asset classes to evaluate financing opportunities and manage overlapping AI exposures at the portfolio level.

AI is a major investment theme, and as new and established companies strive to develop an AI moat, the related capital expenditure (CapEx) will continue to be of historic proportions. As AI drives need for computing capacity to deploy more sophisticated AI models, and the supporting infrastructure from power to cooling, investors are starting to focus on how the substantial capital commitments will translate into sustainable earnings and attractive returns on investment.

Anthropic's IPO filing offers a window into this dynamic. Anthropic reported $16.3 billion in revenue during the first half of 2026, nearly 14 times the amount generated a year earlier. Yet it also incurred nearly $9 billion in computing and infrastructure expenses over the same period, illustrating the capital intensity of developing and operating advanced AI models. ¹

The investment requirements extend well beyond current operations. Anthropic has outlined nearly $800 billion in spending commitments over the next decade, and expectations for its potential IPO valuation continue to fluctuate between $1.2 trillion to $1.75 trillion.² The rapid scaling of high growth companies in private markets highlights the growth ambitions of leading AI developers and the financial resources required to support them.

The broader question for investors is how and where these investments will generate return. Growing demand for computing power creates opportunities across data centers, power generation, networking, and industrial infrastructure, but the benefits may vary across businesses. Capital efficiency, pricing power, and the ability to generate sustainable cash flows are becoming more important as the investment cycle matures.

Active managers are already positioning for this broader opportunity. While earlier investments focused on hyperscalers, semiconductor leaders, and memory producers, managers are expanding into electricity generation, grid infrastructure, cooling systems, networking equipment, and data centers. These businesses are critical to AI deployment and offer investors exposure beyond the largest technology companies.

The scale of AI infrastructure investment reflects the physical requirements of developing and deploying increasingly sophisticated technology. Computing capacity depends on advanced processors, high-speed networking, cooling systems, and reliable electricity. As AI adoption expands, companies must invest in new facilities and upgrade existing infrastructure to support growing demand.

The spending estimates illustrate the magnitude of this investment cycle. Hyperscaler CapEx are projected to reach approximately $697 billion in 2026, while annual AI infrastructure spending could grow from roughly $765 billion currently to $1.6 trillion by 2031. ³

Individual company spending plans reinforce these projections. Alphabet's updated 2026 CapEx guidance of $195 billion to $205 billion is double its $91.5 billion spending in 2025. Amazon has also indicated approximately $220 billion in planned CapEx for 2026, with all hyperscalers continuing to revise upwards their estimates (see figure 1 below).

These commitments create significant financing needs, even for companies with substantial operating cash flow. Hyperscalers can fund considerable investment internally, but the size and duration of the buildout also create incentives to access external capital, preserve financial flexibility, and share project-level risks.

The financing need extends beyond the largest technology companies. Data center developers, power providers, and infrastructure operators also require capital to expand capacity.  Their financing options and balance sheet strength vary considerably, while higher borrowing costs and energy prices can influence project economics. For investors, the ability to generate sustainable returns on capital is becoming as important as the scale of infrastructure spending.

Source: Russell Investments; 2026 CapEx figures reflect the latest company guidance; FCF estimates combine reported cash flows through the latest quarter with assumed Q4 cash flow and spending.

The scale of infrastructure investment is changing how companies raise capital. Corporate cash flow and equity remain important funding sources, particularly for established hyperscalers. However, the capital requirements of data centers, computing equipment, and energy infrastructure are expanding the role of public debt and private credit.

One indication of this shift is the growing use of bond markets. Recent estimates indicate that up to one-third of AI capex funding is being financed in bond markets, ⁴ as external borrowing has become a significant component of the investment cycle.

Private credit is also playing a larger role in financing AI infrastructure. Outstanding private credit to AI-related companies has reportedly grown to more than $200 billion, while projections suggest an additional $800 billion in data center financing could come from private credit over the next two years.⁵

For investors, financing costs are an important consideration. Higher bond yields can increase borrowing costs, while a slowdown in AI capital spending could reduce financing demand. These dynamics reinforce the importance of assessing borrower fundamentals and the durability of infrastructure investment.

Source: Russell Investments, IBES estimates. As of September 2026. Hyperscalers include Alphabet, Amazon, Meta, Microsoft, and Oracle. For illustrative purposes only

As financing needs expand, companies are turning to more specialized structures to fund AI infrastructure. CoreWeave's $7.5 billion GPU-backed debt facility illustrates how computing equipment and customer contracts can support secured borrowing. Meanwhile, Meta's Hyperion data center project combines an equity joint venture with approximately $27 billion in external debt financing.

Public bonds offer exposure to corporate borrowers, while private credit and asset-backed transactions can provide financing tied to specific infrastructure projects, equipment, or contractual cash flows. Each carries different considerations around collateral quality, repayment capacity, liquidity, and refinancing risk.

For investors, this creates a broader set of potential return drivers. Equity investments participate in business growth, while debt investments depend more directly on borrowers' ability to service obligations. Infrastructure and structured finance investments may offer exposure to contracted cash flows or specific assets, and require diligent assessment of project economics and execution risk.

As AI infrastructure investment expands, investors face a broader portfolio construction challenge. Exposure to AI growth may come through newly public companies, established infrastructure providers, or the debt and structured financing supporting the buildout. Each requires different expertise to assess valuations, credit quality, structure of the investment, and project economics.

A multi-manager approach combines these specialized capabilities while providing a coordinated view of portfolio risk. As investors place greater emphasis on capital efficiency and sustainable returns, specialized managers can help distinguish companies and financing structures with durable cash flows from those more vulnerable to rising costs or weaker demand. Combining specialized manager selection with portfolio-level oversight can help investors identify opportunities while managing concentration, liquidity, and overlapping risks.

¹ https://www.ft.com/content/f6ecfecd-d797-4d2c-b427-d333297949cb?sharetype=blocked&syn-25a6b1a6=1
² https://www.ft.com/content/f6ecfecd-d797-4d2c-b427-d333297949cb?sharetype=blocked&syn-25a6b1a6=1
³ J.P. Morgan, Goldman Sachs
⁴ https://finance.yahoo.com/markets/stocks/articles/wells-fargo-says-ai-spending-134523062.html; https://www.goldmansachs.com/insights/goldman-sachs-exchanges/how-ai-debt-is-reshaping-the-credit-market; https://am.jpmorgan.com/wr/en/asset-management/institutional/insights/market-insights/market-updates/on-the-minds-of-investors/can-credit-markets-absorb-the-ai-buildout/
⁵ Morgan Stanley