This article reads Alphabet’s latest earnings through a political-economy lens, viewing artificial intelligence not merely as a technological breakthrough but as an infrastructure-heavy sector where capital, scale and control increasingly shape outcomes. The aim is to situate Big Tech’s AI strategies within broader questions of power, state capacity and global asymmetry, particularly from an Indian perspective.
Google’s Quarterly Results Bedazzle the World—but Not the Stock Markets
Alphabet’s latest earnings mark more than a strong financial quarter; they signal a decisive shift in the political economy of artificial intelligence. Google has moved from being merely “AI-exposed” to unapologetically AI-centric, using its balance sheet to secure control over compute, data flows and distribution at global scale. What is unfolding is not just a race to build better models, but a contest over who owns the infrastructure of intelligence itself—and who, by extension, will shape the rules under which states, firms and societies adapt.
Google’s blockbuster quarter
Alphabet closed 2025 with a record quarter. Revenue came in at roughly $113.8 billion (up about 18% year-on-year), while net income reached approximately $34.5 billion (up nearly 30%). Full-year revenue crossed the $400-billion mark for the first time. Earnings per share were around $2.82, roughly 7% ahead of consensus, while revenue exceeded expectations by about 2–3%. This was not just a “good” print, but a genuine beat on both the top and bottom line.
The engine room was Google Services combined with an AI-turbocharged Cloud business—precisely the combination that allows Alphabet to fund capital-intensive bets internally, without dependence on volatile external capital markets.
“Google Search & Other” revenue reached about $63 billion in Q4 alone, growing roughly 17% year-on-year. This indicates that AI Overviews and related enhancements are, so far, additive rather than cannibalising the advertising machine. YouTube and subscription revenues continued to grow at healthy double-digit rates, reinforcing a diversified services base that throws off steady cash flows.
Google Cloud, however, was the standout. Quarterly revenue accelerated to around $17–18 billion, with growth estimated in the mid-40s percent year-on-year. Operating income more than doubled, driven by demand for Gemini, AI infrastructure, and enterprise AI workloads. In political-economy terms, Cloud has evolved from a challenger business into a strategic lever—one that converts capital expenditure into long-term institutional dependence.
On the cost side, management explicitly signalled much higher capital expenditure. Guidance for 2026 points to roughly $175–185 billion of capex, with AI data centres and custom chips (TPUs) absorbing the bulk of this spending. A $2.1-billion Waymo charge and rising depreciation weighed on operating income, but these were more than offset by revenue growth and continuing efficiency gains. Alphabet has made it clear that near-term margin volatility is an acceptable price for long-term infrastructure dominance.
Investors’ initial reaction was muted but positive, reflecting a familiar tension: extraordinary operational performance balanced against the reality that AI, at scale, is a balance-sheet business before it becomes a software one.
Anthropic: Google’s strategic hedge and profit lever
A crucial but still under-appreciated pillar of Alphabet’s AI strategy sits off the P&L: its stake in Anthropic.
Court filings and subsequent reporting indicate that Google has invested more than $3 billion in Anthropic, giving it roughly a 14% equity stake, with additional convertible financing taking its effective economic exposure into the mid-teens. As Anthropic’s valuation has surged—including a $13-billion Series F at a reported $183-billion post-money valuation—Alphabet has booked multi-billion-dollar unrealised gains, including approximately $10.7 billion as Anthropic’s value was marked up.
This arrangement matters in three ways that go well beyond technology.
First, financially, Anthropic has already become a material contributor to Alphabet’s reported profits via fair-value gains, supplementing operating income from advertising and Cloud. Second, strategically, Alphabet occupies a dual role: Gemini and Claude compete at the model layer, while Anthropic simultaneously remains a major customer of Google Cloud. Alphabet earns infrastructure rents regardless of which model proves superior in specific use cases.
Third—and most importantly—this structure functions as a risk-sharing mechanism. If Claude outperforms Gemini in safety-sensitive or regulated sectors, Alphabet still benefits through equity appreciation and cloud contracts. In effect, Google has diversified model risk while consolidating control over the infrastructure layer.
Alphabet is now structurally long two major model ecosystems—its own Gemini family and Anthropic’s Claude—while sitting beneath both as a provider of compute, storage and networking. This is not hedging in the narrow financial sense; it is power consolidation through capital placement.
The AI landscape: Google versus Nvidia, OpenAI/Microsoft, xAI and others
Viewed from a distance, the AI race rests on four structural pillars: chips, models, infrastructure, and distribution. Alphabet’s latest quarter underscores that it is one of the very few firms with meaningful exposure to all four.
Nvidia remains the archetypal “picks and shovels” supplier, dominating training and inference silicon and benefiting from hyperscaler capex—including Google’s. Alphabet differs in that it is increasingly substituting in-house TPUs where feasible, an attempt to internalise value and reduce dependency, even as it continues to rely on Nvidia for parts of its stack.
OpenAI, alongside Microsoft, continues to define the frontier of general-purpose models and enjoys formidable enterprise distribution via Office, Azure and GitHub. Its recent moves towards open-weight releases suggest a bid to shape the broader ecosystem rather than merely monetise exclusivity.
Anthropic, with Claude, has carved out a strong position as a “safe, steerable, enterprise-friendly” model provider and is now valued in the high hundreds of billions. Google’s equity stake and cloud alignment give Alphabet a rare position: it competes at the model layer while extracting value as a financial investor and infrastructure landlord.
xAI, now folded into SpaceX in a roughly $1.25-trillion private combination, represents a different model altogether: vertically integrated, data-rich, and politically idiosyncratic. Its planned IPO will test whether capital markets reward this alternative form of AI empire.
Against this backdrop, Alphabet stands out as a demand and distribution giant, a rapidly scaling infrastructure provider, and a model builder combined with a financial investor. This makes Google less a single-bet AI company and more a multi-pole AI conglomerate, closer in structure to a utility-cum-platform than a conventional technology firm.
India: why this quarter matters
For India, Alphabet’s AI turn is not merely a Wall Street story; it has direct implications for infrastructure, regulation and state capacity.
In October 2025, Google announced a $15-billion AI data-centre hub in Visakhapatnam, to be built over 2026–2030. It will be Google’s largest AI facility outside the United States, designed as a 1-gigawatt campus expandable to multiple gigawatts. Developed with partners such as AdaniConneX and Airtel, the project will host TPU and GPU compute for large-scale training and inference, tied into subsea cables and national fibre networks.
Investor commentary explicitly links this India build-out to the same capex cycle underpinning Alphabet’s strong quarter. India is no longer treated as a peripheral market, but as a strategic AI geography within Alphabet’s global footprint.
For Indian policymakers and industry, three consequences follow. First, Visakhapatnam effectively places India on the map as a front-line AI infrastructure state, with local access to high-end compute for startups, banks, hospitals and government projects. Second, it raises regulatory and strategic questions around data localisation, energy security, and bargaining power when a handful of global firms control both chips and data pipelines. Third, it lowers the barrier to building export-grade AI services from India, even as it deepens dependence on foreign-owned infrastructure.
Alphabet’s strong quarter financially underwrites this India hub; the hub, in turn, strengthens the strategic logic of Alphabet’s global capex.
Outlook and concluding perspective
Alphabet has now signalled a deliberate trade-off. It is willing to tolerate near-term margin pressure in order to secure a commanding position in AI compute, models and distribution. It is betting that Gemini, Cloud and its advertising stack will monetise this investment at scale—and that its equity exposure to Anthropic will pay off even if innovation leadership shifts outside the firm.
Relative to Nvidia’s chip dominance and OpenAI/Microsoft’s model-plus-enterprise reach, Google’s latest numbers show that it can finance the AI race from internal cash flows, build proprietary infrastructure, and still participate financially in the success of independent labs. For India, this means being plugged directly into one of the world’s main AI power grids—but on terms largely shaped elsewhere.
If this quarter is any guide, the next phase of AI will be shaped less by a long tail of startups and more by a small number of hyper-scale, multi-platform empires. Alphabet’s results—and its quietly valuable Anthropic stake—confirm that Google is not only in that club, but intent on remaining close to its centre.