Alphabet's Earnings Are the AI Spending Trade's First Big Test This Season

Alphabet reports second-quarter earnings after the closing bell, and this one carries more weight than the usual quarterly ritual. Analysts polled by LSEG expect earnings per share of $2.89 on revenue of $116.9 billion. But the numbers investors care most about this time aren't really about the past quarter at all — they're about whether Google's enormous AI bet is starting to pay for itself.

That bet is not small. FactSet's consensus estimate puts Alphabet's capital expenditures for the year at $187.1 billion, part of an industry-wide spending spree on data centers and chips meant to keep up with demand for AI compute. Google, along with the rest of the biggest tech companies, has been pouring money into this buildout for months. The question hanging over the market now is whether that demand is real enough, and durable enough, to justify the bill. Alphabet's results are the first major data point this earnings season that could start to answer it.

Layered on top of that is a more specific worry: Google's own AI roadmap has hit a snag. The company reportedly pushed back the launch timeline for its next flagship model, Gemini 3.5 Pro, a delay that lands awkwardly given how aggressively rivals are moving. Investors will be listening closely for any explanation or reassurance on the call.

Google did make a smaller AI move just this week, rolling out three cheaper Gemini models on Tuesday, including one built specifically for cybersecurity work. That model is aimed at a growing concern in the industry: as AI systems get more capable, so does their potential for misuse, a worry sharpened by the rise of powerful competing models like Anthropic's Mythos. The cheaper Gemini lineup also speaks to a pricing battle playing out across the AI industry, as businesses balk at rising token costs — the per-use pricing that AI companies charge for running their models — and as cheaper Chinese models add pressure from below.

That pricing fight is exactly why investors want a clearer read on how Google's models stack up against frontier labs like OpenAI and Anthropic. It's no longer enough to have a capable model. Companies need one that's competitive on cost, because that's increasingly where enterprise customers are making their decisions.

Search, YouTube and Cloud Under the Microscope

Beyond the AI narrative, the underlying businesses still have to deliver. Google's core search advertising business is projected to bring in $63.4 billion, according to StreetAccount, up from $54.19 billion a year ago. YouTube advertising is expected to hit $10.8 billion, up from $9.8 billion last year.

Cloud is the number to watch most closely. Analysts expect Google Cloud revenue of $22.3 billion, a sharp jump from $13.62 billion in the same quarter last year. Google's cloud unit has been on a strong growth run recently, and it's the clearest evidence so far that the AI infrastructure spending is translating into actual revenue. Whether that pace can hold, and whether Google is closing the gap on Microsoft Azure and Amazon Web Services, will shape how the market interprets everything else in the report.

Why This Report Carries Extra Weight

Alphabet's results matter well beyond its own stock price. With capex across the sector climbing into the hundreds of billions of dollars, any sign that AI-driven cloud demand is slowing, or that model delays are hurting Google's competitive position, would ripple through how investors think about the entire AI spending cycle. Conversely, strong cloud numbers and a credible answer on the Gemini roadmap could reinforce the case that the spending is justified.

Alphabet's earnings call begins at 4:30 p.m. ET, with the numbers arriving after the market close. Given the size of the capex figures involved and the competitive pressure building around AI model pricing, this report is likely to set the tone for how the rest of Big Tech's AI spending gets judged this earnings season.