Microsoft and Google face $650bn AI spending test as sell-off opens fresh entry point

Wedbush says hyperscalers are still in the “early innings” of an AI infrastructure boom, with $650 billion in 2026 capex set to reshape cloud economics and reward scale players despite near-term margin pressure.

AI capex surge reframes the sell-off

A sharp pullback in megacap tech has created what Wedbush calls an opportunity to focus on the structural winners of the AI buildout: Microsoft and Alphabet.

In an industry note, the broker argues that the hyperscaler market remains in the early stages of a structural AI-driven infrastructure expansion.

More than $650 billion in AI capital expenditure has been committed for 2026 by Microsoft Corp (NASDAQ:MSFT), Alphabet Inc (NASDAQ:GOOG), Amazon.com Inc (NASDAQ:AMZN) and Meta Platforms Inc (NASDAQ:META, XETRA:FB2A, SIX:FB), signalling what it describes as an inflexion point in demand.

The core thesis is simple. Hyperscalers are no longer just adding generic cloud capacity. They are redesigning platforms around AI-first workloads that carry higher switching costs, longer contract durations and deeper customer lock-in.

Much of the incremental spend is flowing into long-lived assets with multi-year utilisation curves. That, Wedbush believes, sets the stage for monetisation well beyond the initial model training phase and into higher-margin inference and enterprise use cases.

Microsoft: Azure growth story “underestimated”

Microsoft is positioned as a core AI winner, with Wedbush arguing that the Street still underestimates the Azure growth trajectory into 2026.

The broker expects AI monetisation to become a larger share of the Redmond group’s revenue mix, supporting both growth and margins over the coming years. A key plank of the strategy is embedding advanced AI capabilities across the Microsoft 365 ecosystem and leveraging its vast enterprise install base for distribution.

While Microsoft’s close relationship with OpenAI has been a recent overhang, Wedbush notes that roughly 45% of a $625 billion backlog represents risk exposure.

However, a mooted $100 billion OpenAI funding round at a valuation above $800 billion is seen as reinforcing stability and confidence in that backlog conversion.

At the same time, Wedbush highlights 30% growth in Microsoft’s non-OpenAI backlog, led by new Anthropic commitments, as an additional revenue driver .

The call is clear: 2026 could mark an Azure inflection year.

Alphabet: heavy spend, rising cloud inflexion

For Alphabet, the headline figure is capex of $175 billion to $185 billion in 2026, well ahead of expectations .

The scale of spending is likely to pressure margins in the near term. But Wedbush believes the investment underpins accelerating adoption across AI infrastructure, AI solutions and core Google Cloud Platform products.

Management is already seeing tangible results across Google Cloud and advertising, with AI integration contributing to acceleration in core search and a meaningful inflexion in cloud revenue growth and margins.

Improvements in Gemini models and rising enterprise uptake of AI-native workloads further strengthen Alphabet’s positioning as a long-term AI beneficiary, according to the note.

Free cash flow noise versus long-term control

Wedbush acknowledges that free cash flow optics will remain noisy as spending peaks. But its view is that the platforms investing early and at scale are best placed to capture durable share, pricing power and ecosystem control as AI workloads mature.

In other words, the sell-off may reflect short-term discomfort with capex and margin compression. The broker’s stance is that the real story is long-duration infrastructure leverage and monetisation still to come.

For investors willing to look past the 2026 spending bulge, Microsoft and Alphabet are framed not as overextended spenders, but as architects of the next cloud profit cycle.