Is it time to reassess your bets on the AI boom?
Investors in US technology stocks face a critical moment of reassessment as questions mount over whether the artificial intelligence boom can justify the unprecedented sums being poured into it.
Semiconductor giant Nvidia delivered blockbuster results this week, with second-quarter revenue reaching $96.2 billion—more than double the previous year and comfortably beating Wall Street expectations of $92.2 billion. The company's Data Center division, which powers AI infrastructure, generated $89 billion in revenue, up 117% year-on-year and exceeding analyst forecasts of $85 billion.
Chief executive Jensen Huang struck a confident tone, declaring that artificial intelligence has reached its inflection point.
It's doing useful work. Its tokens—units of data—are productive and profitable. Now, compute is revenue.
Trillion-Dollar Spending Spree
Yet beneath the celebratory headlines lies deepening concern about the economics of AI. The four largest US hyperscalers—Microsoft, Amazon, Alphabet, and Meta—are projected to spend approximately $725 billion on AI infrastructure capital expenditures in 2026 alone, a 77% increase from $410 billion in 2025. Looking further ahead, UBS estimates these companies could spend $4.1 trillion from 2026 through 2028, more than triple the $1.3 trillion deployed over the previous six years.
The scale of this investment is staggering. Amazon, Alphabet, and Microsoft are collectively spending approximately 102% of their cloud revenue on capital expenditures in 2026, effectively recycling all cloud income back into AI infrastructure.
David Coombs of investment firm Rathbones calls these
bewildering timesfor technology investors. Even as he acknowledges that AI
is real and growing exponentially, and changing lives,he warns that US bond yields exceeding 5% would spell trouble for Big Tech companies carrying prodigious debt loads to fund their AI ambitions.
Circular Financing Fears
Adding to investor anxiety is Nvidia's announcement of a partnership with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to establish financing platforms targeting more than $500 billion in third-party capital for AI infrastructure buildout. While the arrangement is designed to help hyperscalers secure funding, it has raised concerns about "circular financing"—where a supplier provides financing to its own customers, creating potential systemic risk if any major player encounters financial difficulties.
The spectre of a "debasement trade" has emerged, with investors increasingly favouring gold and bitcoin over technology stocks. This shift has been triggered by geopolitical tensions, including Donald Trump's escalating confrontation with Iran's trading partners, notably China, and Treasury Secretary Scott Bessent's controversial intervention in the bond market. Bessent has doubled Treasury buybacks to at least $4 billion in an effort to lower long-term yields, a move that has drawn criticism even from Wall Street veterans like hedge fund manager Stanley Druckenmiller.
Community Pushback
Opposition is also growing to data centre construction itself. These facilities, packed with servers and cabling, consume vast quantities of energy and water, prompting resistance from local communities in the United States. This pushback could spread internationally, potentially slowing the infrastructure buildout on which AI development depends.
Investment Strategies Under Review
Despite these challenges, many analysts remain optimistic about Big Tech's prospects. Alphabet, Amazon, Apple, Microsoft, Meta, and Oracle continue to receive "buy" ratings from major brokerages, though some investors avoid Meta on ethical grounds following its settlement of a US court case over the addictive nature of its services for young users.
Matt Britzman of broker Hargreaves Lansdown argues that Meta's $1.5 trillion market valuation already reflects legal risks and AI spending concerns. Nvidia, despite facing competition from Chinese chip manufacturers and in-house chip development by Alphabet and Amazon, also retains "buy" recommendations. The company has guided third-quarter revenue to $108 billion, though notably this outlook assumes no Data Center revenue from China, reflecting the impact of US export restrictions.
Investment professionals emphasise the need for diversification across what Coombs calls
the AI value chain.This means backing companies involved in every aspect of AI development, from chip manufacturers and data centre operators to energy providers supplying the electricity that powers these facilities.
Beyond US names like Alphabet, Amazon, chip software group Cadence, data-centre operator Equinix, and renewable energy specialist WEC, investors are looking at international opportunities. These include Dutch chip-making equipment manufacturer ASML, Taiwanese electronics giant Delta Electronics, and Taiwan Semiconductor Manufacturing Company.
The artificial intelligence revolution may indeed be transforming the global economy, but investors can no longer rely on the same strategies that generated spectacular returns over the past four years. As market volatility increases and questions about returns on AI investment intensify, portfolio reassessment has become essential—even for those who remain believers in the technology's long-term potential.
