Google’s AI spending pushes quarterly free cash flow below zero for the first time

Lead

Google is still one of the most profitable companies in technology, but the economics of the AI race are beginning to show up in a striking way. According to Ars Technica’s coverage of Google’s second-quarter 2026 financial results, the company posted $119.8 billion in revenue, comfortably ahead of analyst expectations. Yet the same report also revealed a milestone investors are unlikely to ignore: Google recorded negative quarterly free cash flow for the first time since going public.

The issue is not a collapse in demand. It is the speed and scale of AI infrastructure spending. Data centers, model training systems, inference capacity, and custom chips are turning the AI boom into a capital-intensive contest.

Key points

• Revenue remains very strong. Search generated $63.3 billion, making it Google’s largest business by far. Google Cloud brought in $24.8 billion, up 23.8 percent from the first quarter, signaling strong demand for cloud and AI services.
• Other major businesses kept growing. Subscriptions, platforms, and devices contributed $12.9 billion, while YouTube ads generated $11.1 billion. YouTube ad revenue rose by more than 12 percent from the prior quarter.
• Free cash flow turned negative. After excluding non-cash investment gains, Google’s operating cash flow for the quarter was about $39.1 billion. But the company spent $44.9 billion expanding its AI footprint, leaving free cash flow at roughly -$5.8 billion.
• Capex guidance moved higher again. Google had previously told investors to expect $180 billion to $190 billion in capital expenditures for 2026. It now says infrastructure spending could reach as much as $205 billion, far above the $91 billion spent in 2025.
• The market reaction was cautious. Google’s share price fell about 4.5 percent overnight after the update and continued trending lower, reflecting investor concern over the cost of the AI buildout.

Why it matters

Negative free cash flow does not mean Google is unprofitable. The company remains highly profitable and holds more than $100 billion in cash and investments. But free cash flow is a crucial measure because it shows how much cash a business has left after funding the investments needed to operate and grow. For a company long associated with massive cash generation, dipping below zero is symbolically important.

The shift also illustrates how generative AI is changing the profile of Big Tech. Investors once valued these firms for software-like margins and dependable cash generation. Now, the race to build and operate AI systems is pulling them deeper into a heavy-infrastructure model.

Google is better positioned than most rivals. Its advertising business remains huge, Google Cloud is growing quickly, and the company designs its own AI chips. Ars Technica notes that the latest Tensor 8i and 8t chips are intended to be more efficient for AI data centers. Still, even these advantages were not enough to offset the quarter’s infrastructure bill.

The next test is whether spending can become revenue at sufficient scale. Google has delayed the release of its Gemini 3.5 Pro model, which it says remains in testing with a small group of partners. Reports cited in the article suggest Google is under pressure to keep pace with GPT 5.6 and Claude Mythos, while also dealing with resignations among senior AI researchers.

For now, Google’s message is that elevated AI capex is not a temporary spike but the new normal. Investors may accept that if AI products and cloud demand continue to grow. If not, the market’s patience with the industry’s spending boom could become much thinner.

Source: Ars Technica AI