AI Infrastructure Spending Faces Investor Scrutiny as Capital Allocation Risks Rise

The rush to build AI capability has turned into a massive fiscal commitment across the cloud and data‑center sector. In the most recent quarter, Microsoft disclosed $30.9 billion in capital expenditures, up more than 80% from a year earlier, while Alphabet reported $35.7 billion, also roughly double its prior‑year spend. Amazon’s quarterly outlay hit $44.2 billion, which extrapolates to an annualized pace of about $175 billion – a figure that dwarfs the company’s free cash flow, now down to $1.2 billion. Meta raised its 2026 capex outlook to between $125 billion and $145 billion, further inflating the total spend among the four largest cloud providers to close to $500 billion. When mid‑tier players such as Oracle and emerging GPU farms are added, industry estimates approach the $1 trillion mark cited by market commentator Peter Schiff.

Investors should note that the bulk of this spending flows to hardware manufacturers. Nvidia posted first‑quarter FY27 revenue of $81.6 billion, an 85% jump year over year, driven largely by data‑center sales that grew nearly 200%. The company’s market valuation now exceeds $5 trillion, reflecting investors’ belief in its central role in AI infrastructure. Micron Technology saw a similar surge after it secured high‑bandwidth memory contracts for GPUs, propelling the stock up more than 225% year to date and delivering a near‑tenfold gain over the past twelve months.

While revenue growth appears robust on paper, several analysts question whether the scale of spending is sustainable. Schiff argues that the equipment being purchased – high‑performance GPUs such as Nvidia’s H100 – may become obsolete within five or six years, far shorter than traditional infrastructure cycles like highways that last decades. If depreciation accelerates faster than anticipated, companies could see a sharp erosion in free cash flow and be forced to reallocate resources away from other strategic initiatives.

The practical impact of this potential misallocation is already visible. Microsoft recently limited access to Anthropic’s Claude Code in favor of its own GitHub Copilot CLI, suggesting that internal AI tools are not yet delivering the expected productivity gains. Uber reportedly exhausted its entire 2026 AI budget by April after deploying Claude Code across thousands of engineers, prompting a reassessment of cost assumptions. These examples illustrate that early‑stage AI tooling can be more expensive and less efficient than projected, raising doubts about the return on capital for large‑scale deployments.

From a market performance standpoint, the companies with the heaviest capex relative to earnings have underperformed their peers. Microsoft’s stock is down roughly 10% year‑to‑date, while Meta has slipped 2.3%. In contrast, Alphabet’s shares are up about 22%, buoyed by a $460 billion cloud backlog that translates spend into booked revenue more reliably. Amazon’s AWS segment posted a 28% growth rate and the stock is up 21% year‑to‑date, reflecting investors’ confidence that its cloud earnings can absorb the massive investment.

The divergence in viewpoints underscores the uncertainty surrounding AI’s financial upside. Microsoft CEO Satya Nadella cites an annual AI revenue run rate of $37 billion – a 123% increase from the previous year – while market sentiment on platforms like Polymarket assigns an 80% probability that Microsoft’s valuation will exceed the combined worth of OpenAI and Anthropic by year‑end. If hardware depreciation follows Schiff’s timeline but revenue growth aligns with Nadella’s projections, investors may see a mixed outcome where some firms benefit from early mover advantage while others face margin compression.

For investors holding exposure to these tech titans, the key considerations are the pace at which capex translates into sustainable cash flow and whether the industry can avoid a wave of stranded assets. Companies that demonstrate disciplined spending, clear pathways to monetizing AI services, and resilience in free cash flow generation are likely to outperform. Conversely, firms that continue to fund hardware without evident revenue traction may see their valuations pressured as the market recalibrates expectations for AI‑driven growth.