AI in Business: Why Investors Doubt the Spending

Amid investor doubts about the sustainability of business spending on AI, semiconductor stocks fell in the first week of July 2026, as investors question whether large technology companies can keep up their current pace of AI infrastructure investment after 2026.

Why the market doubts the sustainability of AI spending

Investors are increasingly focused on one key question: will the growth in AI infrastructure investment hold in 2027 and beyond? That uncertainty became the main source of pressure on the shares of leading semiconductor makers in early July 2026.

According to Mike Shepard, senior editor at Bloomberg Technology & Strategic Industries, big tech companies — Microsoft, Google, Meta, Amazon — continue to publicly confirm large-scale plans to build data centers and GPU clusters. The market, however, has reacted with caution: whether this spending will convert into comparable revenue in the medium term remains unanswered.

Analysts point to several key factors weighing on the sector:

• The cyclical nature of the chip market — even during technological supercycles, semiconductor stocks undergo regular corrections of tens of percent
• Demand concentrated among four or five mega-customers creates a structural dependence on their individual capital spending decisions
• Uncertain ROI on AI investments: most companies have not yet converted multibillion-dollar spending into a comparable increase in revenue for shareholders
• Possible saturation of the GPU accelerator market: some analysts believe large customers will have largely completed their baseline purchases of computing capacity by the end of 2026

Still, the correction reflects not a break in the AI trend but a market repricing of demand growth speed. Fundamentally, the need for AI infrastructure — data centers, accelerators, specialized memory chips — remains structurally durable over a horizon of several years.

Why SK Hynix is entering the US market

Alongside the sector correction, South Korea's SK Hynix, one of the world's leading makers of AI memory chips, announced plans to list on US exchanges through American depositary receipts (ADRs).

ADRs would let US retail and institutional investors buy SK Hynix shares directly in dollars through ordinary brokerage accounts, without having to trade on the Korea Exchange (KRX). For the company, an ADR listing means a broader investor base and direct access to the world's largest financial market.

SK Hynix holds a strategic place in the AI supply chain: it makes HBM (High Bandwidth Memory) chips, a critical component for NVIDIA's H100 and H200 accelerators and AMD's MI300X. HBM provides the bandwidth needed to train the largest language models and run AI agents in real time. Without HBM, building truly high-performance AI clusters is physically impossible — and SK Hynix is one of the world's top three producers of this component.

"SK Hynix's ADR will give American investors easier access to one of the world's leading suppliers of AI memory chips,"

Mike Shepard explained on Bloomberg This Weekend.

The US listing serves a dual purpose: raising capital to fund expanded production of next-generation HBM chips and raising brand recognition among large international investment funds.

What this means

The semiconductor correction is a normal market repricing of growth rates, not a reversal of the AI cycle. Investors are not waiting for AI spending to stop; they want to know when the giant infrastructure investments will start generating comparable profit. Watch the quarterly capex reports from Microsoft, Google and Meta — they set the tone for the entire semiconductor sector. Meanwhile, SK Hynix is timing its move well: an ADR listing positions the company as an accessible vehicle for betting on AI infrastructure just as American investors look for diversification within the AI segment.

*Meta has been designated an extremist organization and is banned in Russia.

Why are chipmaker shares falling?

Investors question whether large technology companies can sustain their current pace of AI infrastructure spending after 2026, despite ongoing public commitments to large-scale capital investment.