Wall Street Thinks AI Is Slowing. Wall Street Is Wrong

Rich Duprey

Wed, July 8, 2026 astatine 10:34 AM CDT 4 min read

Quick Read

• SemiAnalysis projects $11.1 trillion successful cumulative AI infrastructure spending done 2029, with yearly concern topping $2 trillion by 2028 and inactive accelerating.

• AI-related indebtedness backed by GPU contracts and datacenter leases could scope $7.1 trillion by 2029, making it 2nd lone to the U.S. owe market.

• Nvidia captures $0.57 of each hyperscaler AI dollar spent, portion TSMC, Micron, and spot instrumentality makers each clasp captious proviso concatenation positions.

• Don't wait: the expert who called NVIDIA successful 2010 conscionable revealed his apical 10 AI stocks. See the afloat database FREE now.

The communicative astir artificial quality has shifted respective times implicit the past year. Investors person disquieted astir stretched valuations, slowing unreality spending, and whether businesses volition make capable instrumentality connected their concern to warrant the billions pouring into AI infrastructure. Yet the latest semipermanent forecasts suggest the concern rhythm is inactive successful its aboriginal innings. 

According to probe steadfast SemiAnalysis, AI infrastructure spending isn't approaching a highest -- it's accelerating. More importantly, the wealth won't halt with chipmakers. It volition ripple crossed the full semiconductor proviso chain, creating opportunities for companies that manufacture everything from representation chips to the instrumentality needed to physique them.

AI Spending Is Shifting Into a Higher Gear

SemiAnalysis projects cumulative AI IT and datacenter superior expenditures volition scope astir $11.1 trillion betwixt 2024 and 2029, with yearly spending topping $2 trillion by 2028. Instead of flattening out, yearly concern is expected to ascent astir each twelvemonth passim the forecast period.

That forecast reflects much than optimistic projections. Hyperscalers proceed signing multiyear infrastructure contracts portion racing to grow AI capableness accelerated capable to conscionable demand. Even much astonishing is however this enlargement volition beryllium financed.

SemiAnalysis estimates AI-related indebtedness volition scope astir $7.1 trillion by 2029, making it 2nd lone to the U.S. owe market. But alternatively than borrowing against homes, AI infrastructure providers volition get against semipermanent GPU contracts and datacenter lease agreements. Those predictable currency flows go collateral for lenders consenting to concern the adjacent procreation of computing infrastructure.

Don't wait: the expert who called NVIDIA successful 2010 conscionable revealed his apical 10 AI stocks. See the afloat database FREE now.

The effect is efficaciously a caller fiscal plus people built astir AI compute. Granted, that introduces caller risks. If AI adoption oregon monetization disappoints, lenders -- not conscionable shareholders -- would consciousness the effects. But arsenic agelong arsenic request continues expanding, the financing mechanics provides adjacent much substance for infrastructure investment.