Google Posts Its First-Ever Negative Cash Flow Quarter on AI Spending

Google has released its financial results for the second quarter of 2026 and, as usual, the revenue is impressive: total revenue reached $119.8 billion, beating analysts' expectations. Yet the company's shares reacted by falling — by roughly 4.5% overnight. The reason is that spending on artificial intelligence has grown so large that, for the first time in Google's entire public history, its cash flow went negative.

Where the Money Comes From

Search remains the main source — $63.3 billion. Google Cloud brought in $24.8 billion, adding nearly 24% over the quarter, which the company reads as evidence of strong demand for its AI services. Subscriptions, platforms and devices delivered $12.9 billion, and advertising on YouTube $11.1 billion; that last line grew by more than 12%, in part because YouTube's ad spots were made even longer.

Setting aside non-cash income from investments, operating cash flow for the quarter came to about $39.1 billion — 40% more than a year earlier.

Where the Money Goes

The problem is on the spending side. Google had previously promised investors capital expenditures of $180–190 billion for 2026 — already double the $91 billion spent in 2025. Now the bar has been raised to $205 billion. In the second quarter alone, $44.9 billion went into expanding AI infrastructure. The arithmetic is simple: $39.1 billion of income minus $44.9 billion of spending yields a negative free cash flow of $5.8 billion.

The company is still fabulously profitable and holds a reserve of over $100 billion. But free cash flow is the actual money a business uses to finance itself without selling assets or taking on new loans, and its turning negative is an unprecedented event for Google.

The New Normal

Management is making it clear to investors that this is how it will be from now on. It proudly points out that current capital spending is roughly six times the $22 billion spent in 2022, before the AI boom — and next year spending is promised to be higher still. Meanwhile, investors are increasingly wondering whether this race makes sense: across the industry as a whole, AI spending this year could exceed $700 billion.

Google's position is stronger than most: it has a powerful advertising business, a cloud and its own AI chips — the new Tensor 8i and 8t promise greater energy efficiency for data centers. But other signals are cause for concern: the release of the flagship Gemini 3.5 Pro model has been delayed, reportedly because of insufficient progress against GPT 5.6 and Claude Mythos, and a wave of layoffs has swept through the company's leading AI researchers. The coming months could prove decisive for Google's AI strategy.

Source: Ars Technica