Artificial intelligence (AI) has been around for seventy years, but its early adoption was hampered by limited computing power and insufficient investment in supporting infrastructure. Over the past twenty years, the first obstacle has eased considerably. Addressing the second has become a defining investment theme of the decade, as investors seek allocations that balance opportunity with resilience.
For much of the AI boom, the investment story centred on a relatively familiar group of technology companies, from semiconductor manufacturers and cloud providers to firms developing sophisticated AI models. But the opportunity is expanding as capital pours into the physical and financial infrastructure required to support AI at scale, including data centres and supply chains serving the broader ecosystem.
In August, NVIDIA announced that it had signed memorandums of understanding with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to establish independent financing platforms. Together, the proposed partnerships are intended to mobilize more than US$500 billion in third-party capital over time for the build-out of AI compute infrastructure.1 The announcement is notable not simply for its size, but for what it suggests about the evolution of AI as an investment theme.
As AI moves from experimentation toward widespread commercial deployment, it raises a broader question: How much AI exposure may already exist across a portfolio—and where does it actually sit?
Financing the AI Ecosystem
Running progressively powerful AI models requires vast computing capacity. That means more semiconductors and servers, but also data centres, electricity generation and transmission, cooling systems, networking equipment, construction and financing.
The scale of the build-out has significant implications for financing. McKinsey estimates that cumulative global spending on data centre infrastructure could approach US$7 trillion through 2030.2 The International Energy Agency (IEA), meanwhile, has noted that planned data centre investment is becoming too large to rely solely on the balance sheets of technology companies, increasing the role of capital markets and other sources of financing.3
As a result, the AI ecosystem increasingly spans asset classes. Public technology companies remain part of the story, but so do utilities and industrial businesses, infrastructure funds, private equity, private credit and other alternative investments. Together, these enablers of the AI economy signal a growing investment opportunity.
Diversified Assets With Shared Assumptions
Some institutional and family-office investors are already taking a broader approach to assessing AI-related exposure. UBS’s 2026 Global Family Office Report found that 65 percent of surveyed family offices were invested across the AI value chain. At the same time, 37 percent were allocating to power and resources and 37 percent to infrastructure—two areas closely connected to AI’s expansion.4
Taken together, these findings illustrate how AI-related exposure can extend into parts of a portfolio that may not carry an explicit AI label. As a result, a complex portfolio can appear diversified on an asset allocation statement while still being influenced by the same underlying economic theme.
Consider a portfolio that includes shares in major technology companies, an infrastructure fund with data centre exposure, private credit financing AI-related projects and utilities expanding generation capacity to meet rising electricity demand. On paper, these investments sit in four different portfolio categories. Their specific risks and return profiles are not identical, but each depends, to varying degrees, on continued growth in AI-related capital spending.
This does not negate the benefits of a diversified portfolio. It does, however, suggest that diversification by asset class is not necessarily the same as diversification by underlying economic exposure.
Assessing Value Across the AI Value Chain
When wealth also includes privately owned businesses, direct investments or significant real estate holdings, identifying total exposure is only the first step. The next is determining where within the expanding AI ecosystem durable economic value is most likely to accrue.
Rapid investment cycles can produce both winners and excess capacity. AI infrastructure may face constraints ranging from electricity availability and permitting delays to rapidly evolving technology. Assets designed around today’s requirements may need to adapt as computing becomes more efficient or new technologies emerge.
Large amounts of investment don’t guarantee attractive returns across the value chain. Likewise, investing in a large publicly traded technology company is not economically equivalent to committing capital to AI infrastructure. As the opportunity set expands, it becomes important to consider which assumptions underpin each investment and whether the expected return appropriately reflects its liquidity, duration and risk.
A Total-Wealth View of AI
AI is likely to remain an important investment theme, although its long-term impact and investment implications remain uncertain. What is changing is how investors gain exposure to it, both directly and indirectly. Reviewing that exposure across your total wealth with your advisor can help answer a more useful question: Where does your AI exposure sit, what risks connect those investments, and how do they fit together?
As the AI investment ecosystem expands, understanding those connections may matter just as much as identifying the next opportunity.
Sources
