Key Takeaways
- Dan Ives attributes the recent AI-stock sell-off to politics, leadership uncertainty, and market mechanics rather than weakening customer demand.
- Gartner expects worldwide AI spending to reach $2.67 trillion in 2026 and $3.64 trillion in 2027, led by infrastructure investment.
- Nvidia, Microsoft, Alphabet, Broadcom, and AMD remain positioned around measurable infrastructure and enterprise demand, though valuation and execution risks persist.
The recent pressure on artificial intelligence stocks does not signal that customers are abandoning AI projects, according to analyst Dan Ives. His argument is that politics, leadership turmoil, and market mechanics have combined into a "perfect storm" for technology shares, obscuring the continued expansion of underlying demand.
That distinction matters. Share prices can fall because investors reduce exposure, rebalance portfolios, react to political uncertainty, or question corporate leadership. None of those developments necessarily means cloud providers and enterprises have stopped ordering chips, servers, networking equipment, storage, and AI services.
In fact, Ives maintains that the AI build-out remains on course to approach roughly $4 trillion over the next several years. As outlined in a recent report on his AI-stock outlook, the more important signal is spending by customers rather than short-term movement in technology-sector valuations.
Industry spending forecasts remain large enough to support that broad thesis. Gartner forecasts worldwide AI spending of $2.67 trillion in 2026, representing 49.5% year-over-year growth (source). AI infrastructure alone is expected to account for $1.48 trillion. For 2027, Gartner projects total spending of $3.64 trillion, including nearly $1.98 trillion directed toward infrastructure (source) (source).
Those figures cover a broad AI market, so they should not be compared directly with narrower hardware forecasts. IDC estimates that spending on AI server and storage systems will total $497 billion in 2026 and rise to $1.21 trillion by 2030. The lower number largely reflects scope: IDC is measuring specific infrastructure categories, while Gartner's estimate encompasses a wider range of AI spending.
Corporate results add another layer of visibility. Google Cloud revenue reportedly increased 103% to $30.7 billion in Q3 2026, while Alphabet's Gemini application reached 1 billion users. Broadcom, meanwhile, reportedly recorded a 221% year-over-year increase in AI-semiconductor revenue to $16.7 billion in fiscal Q3 2026 and issued fourth-quarter guidance of $21.7 billion.
Nvidia has also raised its growth forecast, reinforcing the view that demand for accelerated computing remains substantial. Microsoft, Alphabet, Broadcom, AMD, and Nvidia all participate in different portions of the AI supply chain, from cloud capacity and foundation-model distribution to processors, networking, and custom silicon.
Could customer demand eventually cool? Certainly. Massive capital programs introduce concrete challenges, including power constraints, stretched supply chains, regulatory friction, and the possibility that enterprise revenue develops more slowly than infrastructure spending. High valuations can also make otherwise healthy companies vulnerable to sharp corrections when expectations shift even slightly.
That said, businesses should separate the market narrative from operating evidence. For technology buyers, the continued build-out could expand access to computing capacity and create more competition among cloud and semiconductor suppliers. It could also bring tougher procurement questions. Buyers increasingly need to assess total operating costs, data controls, model performance, vendor concentration, and governance rather than selecting products mainly because they carry an AI label.
The same filter applies to investors. Infrastructure suppliers and established cloud operators currently offer more visible measures of demand than speculative application-layer businesses with uncertain economics. Revenue growth, backlog, customer adoption, capacity utilization, and cash generation can provide a firmer basis for evaluation than broad enthusiasm alone.
Ives' central point is not that every AI stock will recover or that every vendor will benefit equally. It is narrower and more useful: the sell-off appears disconnected from evidence of collapsing customer demand. With spending forecasts still rising and major suppliers reporting substantial AI-related growth, the sector's immediate challenge looks more like confidence and valuation than the disappearance of its underlying market.
⬇️