2 hours ago
12
The numbers coming out of Big Tech’s AI buildout are getting hard to contextualize. Alphabet, Amazon, Meta, and Microsoft have collectively committed close to $2.4 trillion in long-term AI infrastructure spending, covering data centers, equipment leases, and the staggering energy costs required to keep it all running. That figure, reported by Bloomberg on July 31, 2026, is not a market cap or a valuation. It is a forward spending pledge, money these companies have already promised to spend.
To put it another way: Alphabet alone disclosed $902 billion in purchase commitments and leases. Meta reported nearly $700 billion. These are not projections from hopeful analysts. They are numbers pulled from corporate disclosures.
A $2.7 trillion correction arrived right on schedule
The same companies making these historic spending commitments watched a combined $2.7 trillion in market value evaporate during a June 2026 selloff, driven by concerns over returns on heavy capital expenditures.
To fund all of this, the five largest data-center spenders collectively added roughly $350 billion in debt over the past five years, effectively doubling their long-term debt loads.
For context, global AI sales excluding China reached $25 billion in the first quarter of 2026 across hyperscalers and neoclouds. That is real revenue, but measured against $2.4 trillion in committed spending, the math requires patience that markets do not always have.
Bitcoin miners found a side door into the AI economy
Bitcoin miners — IREN, Hut 8, TeraWulf, and Core Scientific — have all announced multi-year AI and high-performance computing contracts, with the total value of these partnerships reaching approximately $90 billion.
AI revenues are expected to climb from roughly 30% of total income for these miners to approximately 70% by the end of 2026.
After Bitcoin’s most recent halving, mining revenue per block dropped significantly, compressing margins for everyone. Meanwhile, hyperscalers are paying premium rates for GPU-dense computing capacity and have long-term contracts to offer, making AI lease contracts a more predictable revenue stream than Bitcoin price appreciation.
Energy costs, and who pays them, is now a political question
President Trump’s July 2026 “Ratepayer Protection Pledge” reflected a consensus among data-center operators and utilities that the cost of AI-driven power consumption should be borne by corporations, not passed on to residential consumers.
What this means for investors watching both markets
Miners that have secured long-term AI and HPC contracts are effectively de-risking their revenue in a way that traditional Bitcoin-only operations cannot. If AI revenues do reach 70% of total income for companies like Core Scientific and Hut 8, their stock behavior will increasingly correlate with AI infrastructure trends rather than Bitcoin price movements.
The debt accumulation at Big Tech scale creates a sensitivity to interest rates that did not exist when these companies were largely debt-free. If rates stay elevated, the cost of servicing $350 billion in new debt becomes a recurring drag on earnings, which could extend the timeline for any valuation recovery after the June selloff.
Investors in both AI infrastructure equities and mining stocks should watch for state-level utility commission decisions, which will be where the actual cost allocation fights play out. A regulatory ruling that forces data centers to pay higher grid connection fees could shift the economics of any project that relies on cheap power access, which describes virtually every major AI and crypto mining operation currently under development.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
English (US) ·