Alphabet’s latest quarterly results delivered a neat contradiction: revenues are rising fast, yet management is signalling an even more aggressive tilt toward capital spending that has rattled markets. Fourth‑quarter revenue climbed 18% year‑on‑year and the company crossed $400 billion in full‑year sales for the first time, with net income of about $132.2 billion. That financial strength, rather than prompting a return to shareholder‑friendly conservatism, has instead been channelled into an extraordinary forward commitment to AI infrastructure.
For 2026 Alphabet guided roughly $180 billion of capital expenditure — almost double its prior year outlays and some 50% above many analysts’ expectations. On a headline basis that works out to about $500 million per day dedicated to data centres, chips, power and other AI‑related build‑outs. The scale is notable not just for its size but for what it says about the changing nature of the AI competition: performance will be won as much by land, wires and electrons as by algorithms and talent.
The earnings print also showed why Alphabet feels emboldened. Google Search revenue rose 17% year‑on‑year, a sign that generative AI features such as Gemini integrations and an array of product updates are complementing rather than cannibalising the company’s core cash machine. Google Cloud was the standout business, growing nearly 50% in the quarter and delivering an annualised run‑rate north of $70 billion, buoyed by sales of enterprise AI infrastructure and services.
Behind the headlines are operational improvements that explain Alphabet’s confidence. Gemini service unit costs fell by about 78% last year thanks to model optimisation, higher utilisation and hardware efficiency gains, allowing the company to price APIs aggressively — reportedly down to $0.50 per million tokens — and expand adoption. User engagement also accelerated after Gemini 3 and subsequent product launches, with hundreds of millions of new users added over recent quarters and longer, more complex queries becoming routine.
Alphabet is pairing this software progress with heavy investment in hardware and energy. The company’s roughly $4.75 billion acquisition of clean‑energy developer Intersect Power gives it the option to build onsite solar, wind and battery capacity to ease grid constraints and shorten data‑centre permitting cycles. That move highlights an often‑underappreciated bottleneck: in many markets the limiting factor for AI scale is not chips but access to reliable, affordable power and the right sites.
Investors reacted with whiplash. Alphabet’s stock swung sharply in after‑hours trading, reflecting a broader ambivalence: strong topline growth and evidence that AI is monetising, counterbalanced by a staggering near‑term cash burn and the prospect of higher depreciation and lower near‑term earnings growth. Analysts warn that much of the $180 billion will convert into future depreciation charges that could dampen profit growth even if revenue continues to climb.
The strategic picture is stark. Alphabet’s bet further raises the capital and energy barriers to compete in large‑scale AI, favouring integrated incumbents with global data‑centre footprints and chip partnerships. At the same time, it concentrates risk: a winner‑takes‑most outcome would amplify the political and regulatory stakes, and prolonged heavy capex could expose the company to cyclical vulnerabilities if AI monetisation slows.