Foxconn Chairman Says Cloud AI Spending Could Near $1 Trillion in 2027

Foxconn Sees a Trillion-Dollar AI Buildout on the Horizon

Foxconn Chairman Young Liu has delivered one of the boldest signals yet about the scale of the global artificial intelligence buildout. Speaking to shareholders at the company's annual meeting on Friday in New Taipei City, Taiwan, Liu said capital expenditure by major cloud service providers has already surpassed $700 billion this year and could approach $1 trillion in 2027.

"Their capital expenditure is our market," Liu told shareholders, referring to the spending of cloud providers. "It has already reached $700 billion, and their capital expenditure next year is expected to potentially reach $1 trillion. This gives us immense confidence in our future growth momentum."

A Structural Shift, Not a Seasonal Peak

The projection underscores how demand for AI infrastructure has rewired the technology supply chain, eliminating the traditional mid-year seasonal slowdown that once defined the industry's rhythm. For Foxconn, the world's largest contract electronics manufacturer, the spending wave represents a structural shift rather than a cyclical peak. Liu said he was "very optimistic" about the second half of the year.

Foxconn expects its own capital expenditure to grow by roughly 30% this year, up from last year's T$174 billion (about $5.55 billion), as it expands manufacturing capacity for AI servers. The company is also scaling production of co-packaged optics, a data transmission technology critical to next-generation AI systems, with plans to deliver units this year to clients building large-scale infrastructure.

Dell's Raised Outlook Reinforces the Trend

Liu's comments came one day after Dell Technologies raised its annual revenue forecast, projecting around $60 billion in AI server revenue for fiscal year 2027, up from a prior estimate of $50 billion. Dell now expects total revenue between $165 billion and $169 billion for the year ending January 2027, a sharp jump from earlier guidance. The back-to-back announcements from two of the largest players in the AI hardware chain paint a picture of demand that keeps accelerating, echoing Nvidia chief Jensen Huang's view that the buildout is still in its early stages.

A Market Transformed

Research firm TrendForce estimates AI server shipments will grow by more than 28% this year, with total AI training compute across the five largest North American cloud providers expected to rise by over 56% year-on-year. AI inference compute is projected to surge by roughly 122%. Taken together, the figures suggest the AI infrastructure race is reshaping corporate capital allocation on a historic scale, with the supply chain straining to keep pace.