Big Tech is betting billion-dollar big on artificial intelligence (AI).
According to recent quarterly earnings reports, the collective capital expenditure (capex) for hyperscalers Amazon.com Inc., Microsoft Corp., Google parent Alphabet Inc., and Meta Platforms Inc. are projected to reach $725 billion in 2026. This represents a significant upward revision from earlier estimates of $670 billion, signaling that the AI spending cycle is accelerating rather than cooling.
“Amazon, Google, Microsoft, and Meta all understand that the risk of underinvesting in AI Capex is significantly larger than overspending,” said Daniel Newman, CEO of The Futurum Group. “Less than 2% of people are paying for frontier AI models. We are still so early.”
For decades, annual capex in the tens of billions was considered aggressive. Today, that is the baseline.
Indeed, the scale of investment is now driven primarily by the need for massive data centers and high-performance silicon required to run generative AI models, largely supplied by NVIDIA Corp. and Taiwan Semiconductor Manufacturing Company (TSMC), who are benefitting from the “picks and shovels” phase.
For the major hyperscalers, capex as a percentage of operating cash flow is projected to surge to nearly 92% by 2026. This means almost every dollar they earn is being plowed back into NVIDIA chips and power grids.
Amazon leads the pack with a projected $200 billion spending outlay. CEO Andy Jassy remains steadfast in this valuation, citing “seminal opportunities” in AI, robotics, and satellite technology.
Microsoft and Alphabet are locked in a near-tie, with both forecasting approximately $190 billion in investments. Microsoft’s spending represents a 61% surge over 2025 levels.
Meta has raised its outlook to a range of $125 billion to $145 billion, citing rising component costs and infrastructure demands.
While the spending figures are universally high, Wall Street’s reaction has been split. Alphabet has successfully charmed investors; its cloud division reported a 63% revenue jump, providing bottom-line proof that AI investments are already yielding returns.
In contrast, Meta has faced a more cautious reception. Unlike Google, Meta’s AI initiatives have yet to show a direct, massive impact on quarterly profits, leading some analysts to question the sustainability of such a steep spending curve.
Standing in stark contrast to its peers is Apple Inc.
The iPhone maker has faced criticism for a perceived slow start in the AI race, but its financial strategy remains disciplined. Based on a fiscal second quarter expenditure of $4.3 billion, Apple’s annual capex is on track for roughly $13 billion.
This measured entrance suggests Apple – whose longtime CEO, Tim Cook, is stepping down Sept. 1 – is leveraging its existing ecosystem rather than attempting to outbuild the cloud giants in raw data center capacity.
Tech executives are framing these expenditures not as costs, but as essential foundations. Microsoft Chief Financial Officer Amy Hood noted that the company remains “confident in the return on these investments,” pointing to higher demand signals and increased product usage.
As these five titans prepare to funnel nearly three-quarters of a trillion dollars into hardware and infrastructure, the message to the market is clear: the AI revolution will not be cheap, and the barrier to entry has never been higher.
The question remains, however: Will the market remain patient with these spending levels if we don’t see a killer app that brings that 2% of paying users up to 20% by next year?
