Big Tech's $3 Trillion Hidden AI Commitments Spotlight Investment Surge and Market Shift
Analysts estimate that major technology firms carry roughly $3 trillion of AI‑related future commitments, far exceeding reported capex. Goldman Sachs forecasts global AI investment to hit $1 trillion in 2026 and reach 1.4% of world GDP by 2028. The gap has spurred a market focus on financing structures, credit exposure, and selective investment in AI winners across regions.
A Wall Street Journal analysis found that nine large technology companies carry roughly $3 trillion of AI-related commitments that have not yet appeared on their balance sheets, against about $600 billion of reported capital expenditure over the preceding year. The gap is not an accounting trick so much as a timing one: roughly $1.2 trillion is leases on facilities that have not begun operating, and about $1.9 trillion is commitments to purchase chips, equipment, energy and other services.
Individual disclosures show how fast the numbers are moving. GOOGL disclosed $811 billion of purchase and contractual obligations, up from $332 billion three months earlier, and META reported $347 billion of future lease commitments. Goldman Sachs projects the group's reported capex rising from $405 billion in 2025 to roughly $750 billion in 2026 and approaching $1.2 trillion in 2027.
Financing is the part that has started to reprice. Those companies issued $194 billion of investment-grade rated debt in the first half of 2026, on track for roughly $250 billion for the year, or about a third of capex. In August, NVDA signed memoranda with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to build compute financing platforms targeting more than $500 billion of third-party capital for data centers and GPU clusters , moving a growing share of the buildout off the hyperscalers' own balance sheets.
The market narrative is shifting accordingly, from how much is being spent to who converts it into earnings. Reuters notes capex angst fading as investors hunt for the firms best positioned to monetize the spend, and the same rotation is visible across Asia-Pacific allocations.
The risk to watch is credit rather than demand. Off-balance-sheet leases and multi-year purchase obligations are fixed claims that do not flex if AI revenue arrives more slowly than the buildout assumes. Investors may track debt issuance spreads, lease commencement schedules, and disclosed obligation growth quarter over quarter as the cleanest early read on whether the financing structure is straining.
Related Stocks
Powered by SentiSense - Intelligent Market Analysis