Silicon Valley’s $700 Billion Question: Capital Allocation in the AI Build-Out

An executive walking through a hyperscale data center

The artificial intelligence boom is, above all, a capital spending boom. Reporting in November 2026 estimated that firms are pledging about $700 billion in AI infrastructure spending this year. Meta has kept its 2026 capital expenditure guidance at $115 billion to $145 billion. Oracle recorded about $55.7 billion in capital spending for fiscal year 2026, according to TechTarget, while expecting restructuring costs of up to $2.1 billion in severance and related expenses.

The same companies are reducing headcount. Oracle’s workforce fell by about 21,000 people, roughly 13 percent, in fiscal 2026, and Meta cut about 8,000 positions in May. Analysts have noted that some of the savings from layoffs are being redirected to data centers and chips.

A classic corporate finance problem at an unprecedented scale

Every MBA student learns to evaluate investments by comparing expected returns with their cost and risk. The AI build-out tests that discipline. Data centers take years to build and depend on electricity, chip supplies and regulatory approvals. Demand for AI services is growing rapidly, but revenue models are still evolving, and some observers warn of a construction bubble. Companies that build too little risk losing market position; those that build too much risk stranded assets.

These are not abstract questions for the technology sector alone. Banks, utilities, governments and pension funds are financing the build-out. Energy companies are planning generation capacity around data center demand. Countries from the United States to Morocco are competing to host facilities.

What future managers need to understand

We believe MBA programs should equip graduates to evaluate AI investment decisions with rigor:

  • Scenario analysis. AI returns depend on adoption rates, pricing, energy costs and competition, all of which are uncertain. Decisions should be tested across several futures.
  • Total cost of ownership. Hardware is only part of the cost. Power, cooling, talent, depreciation and the opportunity cost of capital must be included.
  • Build, buy or partner. Most organizations will not build data centers. They must decide when to rent cloud capacity, when to use specialized providers and when shared or public infrastructure makes sense.
  • Linking investment to workforce strategy. Funding AI by cutting staff is a choice with consequences for capability, morale and reputation that belong in the investment case.

Where the Global South fits

For much of the world, the question is not whether to spend hundreds of billions, but how to secure affordable access to AI capacity and avoid permanent dependence. Projects such as Morocco’s Nexus AI Factory, shared research computing, and regional cloud investments all represent different answers. Each requires public and private leaders who can negotiate with global technology firms from a position of knowledge.

EUCLID’s MBAs aim to connect these worlds. Our students study the strategies of Silicon Valley’s largest investors alongside the development priorities of the Global South, so that they can make sound capital decisions wherever they lead.

Sources

aiMBA Insight is published by the EUCLID and EULER MBA programs. Explore our MBA in Artificial Intelligence and Digital Transformation and our other specialized MBAs.

More from aiMBA Insight

Lead the AI transition

Specialized online MBAs from EUCLID and EULER, built for professionals in the public and private sectors.