Alphabet: All In on AI

This is the heart of the case. Alphabet's 2025 operating cash flow comfortably exceeds even its enormous 2026 spending plans, which run to roughly $190 billion, more than double the prior year. While OpenAI, Anthropic, and others must raise fresh capital every cycle to stay in the race, Alphabet can self-fund, cut back if the race slows, and still outlast competitors. In June 2026 it chose to top up that firepower with its largest-ever equity raise of around $85 billion, with Berkshire Hathaway taking a tenth of it, a notable vote of confidence from the world's most famous value investor.

For years Google Cloud trailed Amazon and Microsoft. That gap is now closing fast. Cloud revenue passed $20 billion in a single quarter for the first time, growing far quicker than either rival, while its order backlog has roughly doubled to around $460 billion, a sign that demand is outrunning the capacity Google can build. Management says enterprise AI tools are now the main driver. A division once viewed as a costly also-ran has become one of Alphabet's most valuable assets and the clearest proof that its AI spending is translating into revenue.

The loudest fear about Alphabet was that AI answers would gut its search advertising. It hasn't happened. Search revenue is still growing despite the rollout of AI Overviews, which do reduce clicks on traditional blue links. The trade-off is subtle but powerful: richer AI queries give Google a sharper read on what users actually want, supporting more precise, higher-value advertising rather than less. The cash engine that funds everything else is proving far more resilient than the bears expected.

• Quarterly results showing Search still growing: Each quarter advertising holds up further undermines the AI-disruption fear that has capped the rating.

• Converting subscribers into payers: New usage caps on heavy AI users mark the start of turning a huge free base into paying customers.

• Cloud backlog turning into revenue: Converting the roughly $460 billion backlog into recognised sales, while holding margins, would confirm cloud as a durable profit engine.

• Agentic and commerce products: Gemini-powered agents and a new commerce protocol with major partners open paths to AI revenue beyond traditional search ads.

• Waymo commercialisation: A re-rated robotaxi unit with rising rides and real revenue could become a genuine profit centre, with an eventual IPO or spin-off as further upside.

• Distribution moats compounding: Deals placing Gemini on Apple and hundreds of millions of Samsung devices, plus a large India commitment, lock in reach that rivals cannot easily match.

Alphabet plans to spend around $190 billion on infrastructure in 2026. If AI demand disappoints or the competitive picture shifts suddenly, returns on that investment could deteriorate fast. The decision to raise about $85 billion in equity, on top of an earlier bond sale, also raises a fair question: if the most cash-generative company on the planet is tapping outside capital rather than fully self-funding, how sustainable is the entire industry's spending trajectory? Heavy spending shows up as rising depreciation and energy costs, which weigh on margins and cash flow.

A US court has ruled that Google holds an illegal monopoly in search. The 2025 remedies were milder than feared, but the Justice Department and many states are appealing for tougher measures, a separate case over Google's advertising technology is still awaiting judgment, and EU rules add further pressure. With advertising still around three-quarters of total revenue, any adverse structural ruling, or a cap on pricing flexibility, would have an outsized impact. The appeals run well into 2027, keeping uncertainty hanging over the core business.

Two competitive threats sit side by side. AI-native search tools such as Perplexity and OpenAI could, over time, chip away at the habit of "Googling," even though there is no evidence of revenue erosion yet. More subtly, Google's own shift toward answering queries directly, rather than sending users to other websites, risks hollowing out the publisher ecosystem that feeds its index. The logic is that if Google doesn't cannibalise its referral model, a rival will, but the strategy still carries the risk that ad economics suffer as the old search bargain erodes.