Big Tech Plans Up to $740 Billion in 2026 Capex as Healthcare AI Deals Multiply

Four companies are planning to spend roughly the GDP of a mid-sized country on computing infrastructure this year. Alphabet, Amazon, Meta and Microsoft have disclosed or indicated combined 2026 capital spending of about $715 billion to $740 billion.

That figure covers infrastructure and other capital investment. It is not AI spending alone.

The spending plans

Alphabet forecast capital expenditures of $195 billion to $205 billion for 2026. Amazon expects roughly $200 billion.

Microsoft projects about $190 billion in calendar 2026, and about $25 billion of that comes from higher component prices. Meta forecast $130 billion to $145 billion, a range that includes principal payments on finance leases.

Amazon CEO Andy Jassy has pushed back on the idea that the money is a gamble. "We're not investing approximately $200 billion in capex in 2026 on a hunch," he said. Much of Amazon Web Services' spending is expected to be monetized in 2027 and 2028.

Why healthcare

The scale of the target is the draw. U.S. national health spending rose 7.2% in 2024 to $5.3 trillion, or 18% of GDP and $15,474 for every resident.

Hospitals accounted for $1.6347 trillion of that. Physician and clinical services came to $1.1097 trillion.

A Fortune commentary argues that healthcare is where the AI bet gets tested, citing an estimate that close to $1 trillion of U.S. health spending goes to administration. It also points to a clinical workforce that cannot keep up with demand. Its logic: time saved on documentation, scheduling and intake can go toward seeing more patients rather than cutting staff.

The deals

Microsoft and Mayo Clinic announced a collaboration on June 2, 2026, to build a healthcare-specific frontier AI model. It will draw on Mayo's de-identified clinical data and longitudinal insights.

The model is initially meant for use inside Mayo's own clinical environment. Mayo CEO Gianrico Farrugia, M.D., said the clinic "is committed to putting patients first, and we have long believed AI can help transform healthcare." Mustafa Suleyman, CEO of Microsoft AI, said, "Frontier medical intelligence is around the corner."

On March 5, 2026, CVS Health and Google Cloud announced a strategic partnership. Gemini models and Google Cloud's healthcare-data and analytics tools will go into CVS's Health100 platform.

Medicare takes a cut

The catch for providers is that Medicare already prices in efficiency. Productivity adjustments are built into payment updates for hospitals, skilled nursing facilities, home health and other settings.

For 2026, CMS extended that logic to physicians. It cut by 2.5% the work-value and intraservice-time inputs used to price specified physician services that are not billed by time. CMS said the adjustment reflects efficiencies it expects to accrue in how those services are delivered over time.

The Fortune commentary reads this as a clock. Providers that use AI early to lower documentation, intake and coordination costs keep the savings for a while. Those that wait could face the same payment pressure without having captured them.

The adjustment is narrow, though. It applies to specified Medicare physician services and does not automatically reduce all healthcare reimbursement as providers adopt AI.

For taxpayers, the arrangement cuts both ways. If AI really lowers the cost of care, Medicare's pricing formula is the mechanism that passes some of the savings to the public purse. Providers that expected to pocket them have a different view of the same formula.

Where the evidence is thin

AGF Investments, an asset manager, says many AI use cases in healthcare "remain quite speculative." It points to two areas where it sees real benefit now: drug development and diagnostics.

AGF says AI applied to standard CT scans and MRIs is already helping clinicians detect anomalies and lesions that would otherwise go unnoticed. It also says AI is improving surgical robots, though fully autonomous procedures are "a long way" off.

The firm is writing for investors, not as a neutral party. Its view is that precision oncology and life-science tool companies are the sub-sectors to watch. It cites a looming revenue cliff for large drugmakers within three to five years as the pressure driving acquisitions.

None of that shows administrative savings at hospitals or physician practices at the scale the capex numbers imply. The Mayo model has not yet been deployed beyond its initial clinical environment, and Mayo has not said when that will change.

The open question is timing. Amazon says its payoff arrives in 2027 and 2028, and CMS has already built a 2.5% efficiency assumption into 2026 physician pricing. Whether measured savings show up in healthcare before the next round of payment updates will show how much of the bet is real.