- Alphabet raised its full-year capital expenditure guidance to between $195b and $205b, up from the previous range of $180b to $190b.
Alphabet delivered a second quarter that was, by nearly every conventional measure, exceptional. Yet the market’s reaction — shares sliding more than 2 per cent in after-hours trading — tells a more complicated story about what investors now demand from the AI era’s biggest players.
The Google parent reported earnings per share of $9.11 on revenue of $119.8 billion, comfortably exceeding analyst expectations of $116.9 billion. Revenue growth hit 24 per cent year over year, a pace that would thrill most companies of Alphabet’s size. But in a moment defined by soaring AI investment and intensifying scrutiny of those outlays, even a beat-and-raise quarter can land as a disappointment.
The standout figure was Google Cloud, which generated $24.77 billion in revenue — an 82 per cent surge from the $13.6 billion recorded a year earlier. That performance outpaced the $24.56 billion analysts had projected and underscored the degree to which demand for AI infrastructure is reshaping Alphabet’s revenue composition.
Advertising, still the company’s core engine, brought in $81.63 billion against expectations of $81.12 billion, a solid if less dramatic showing.
CEO Sundar Pichai captured the moment succinctly. “Q2 was an amazing quarter, with Alphabet revenues growing 24 per cent year over year and Google Cloud revenues accelerating to 82 per cent growth, driven by demand for AI infrastructure and AI solutions,” he said in a statement accompanying the results.
The cloud acceleration is significant beyond the numbers. It suggests that enterprises are not merely experimenting with AI but are committing real budget to cloud-based AI workloads — a shift that positions Google Cloud as a direct beneficiary of the same AI investment cycle that is pressuring the company’s capital expenditure line.
Spending signal shakes confidence
That pressure became the evening’s focal point. Alphabet raised its full-year capital expenditure guidance to between $195 billion and $205 billion, up from the previous range of $180 billion to $190 billion. Analysts had been expecting roughly $186.4 billion. The magnitude of the increase — up to $25 billion at the high end — caught the market off guard, and shares fell more than 2 per cent on the news.
The reaction reflects a growing tension in how investors process the AI buildout. For much of the past two years, markets rewarded aggressive spending on data centers, chips, and AI infrastructure as evidence of ambition. But the calculus is shifting. With cumulative AI-related capital expenditure now running into the hundreds of billions across the industry, investors are beginning to ask harder questions about when — and whether — those investments will translate into sustained margin expansion.
Alphabet’s remaining performance obligations, a forward-looking metric reflecting contracted revenue yet to be delivered, reached $514 billion, well above the $488.1 billion Wall Street had anticipated. That figure offers some reassurance that demand is real and durable. But it does not fully quiet the question of timing and return.
Competitive landscape
Alphabet’s position in the AI race remains strong but contested. The company, like rivals Microsoft, Amazon, and Meta, faces the dual pressure of investing massively in infrastructure while simultaneously advancing the capabilities of its models. Any stumble on the latter front could give competitors an opening.
A media report suggested that Google has delayed its Gemini 3.5 Pro model over concerns about its capabilities relative to competing systems. Google pushed back firmly. “We’re shipping quickly across a wide range of models while keeping them highly cost-effective for customers,” a spokesperson said.
“We’re currently testing 3.5 Pro, an upgraded Flash model, and other models with partners, and we’re productively engaged with the US government.”
The exchange highlights the delicate position AI leaders occupy: racing to deploy ever-more-capable systems while navigating regulatory engagement, cost discipline, and the reputational stakes of releasing models that underperform expectations.
On the hardware side, Alphabet received a boost earlier in the week when The Information reported that Google is developing a new chip, codenamed Frozen v2, designed to improve efficiency when running Gemini.
The chip would embed components of the model directly, reducing processing time and potentially lowering the cost per inference — a development that, if successful, could address some of the very cost concerns that unnerved investors on Wednesday.