The Money Still Comes From Ads
First, let's follow the money. The vast majority of Alphabet's eye-watering revenue—expected to be around $116 billion for the second quarter of 2026—still flows from its established advertising empire. The Google Search and YouTube ad machines are incredibly
mature and profitable, forming the financial bedrock that allows the company to make enormous bets on the future. While AI features like AI Overviews are being integrated into Search, their primary role so far is defensive: to keep users engaged as search behavior evolves. For now, the core profit engine remains the classic, highly-tuned ad auction model, not the shiny new generative AI tools.
AI Is an Expensive Bet
Developing and running large-scale AI is staggeringly expensive. Alphabet has guided its capital expenditures (CapEx)—the money spent on physical things like data centers and specialized chips—to a range of $180 billion to $190 billion for 2026. This represents a massive increase in spending directed almost entirely at building the infrastructure needed to power models like Gemini and serve AI-hungry cloud customers. These billions are an investment, not a profit. The cost of just one complex AI query is significantly higher than a traditional search. This dynamic positions AI as a massive cost center today, with the hope of it becoming a profit center tomorrow. For now, free cash flow is feeling the pressure from this spending surge.
The Search for a Business Model
The central question Google is still trying to answer is how to directly monetize consumer-facing AI. While features like AI Overviews are live, it's not yet clear if they increase ad revenue or just costs. Google is experimenting with new ad formats that embed promotions directly inside AI-generated answers, but this is a delicate balance. The other path is subscriptions. Products like Google One, which bundle storage with access to more powerful Gemini models, are a growing source of revenue, but they are a small fraction of the advertising juggernaut. Unlike its ad business, where the monetization is clear, the playbook for making money from generative AI in its core products is still being written.
The Exception: Google Cloud
There is one area where the AI-profitability picture is much brighter: Google Cloud. This division is a primary growth engine for Alphabet, with revenue expected to soar by over 60% year-over-year. Google Cloud makes money by selling AI infrastructure and tools directly to other businesses. Companies building their own AI products are renting Google's powerful servers and models, turning Google's massive CapEx into direct revenue. Major deals, including a significant commitment from AI company Anthropic, have driven a cloud backlog of over $460 billion. This enterprise-focused strategy is clearly working and represents the most mature part of Google's AI monetization plan, even if it's less visible to the average consumer.
