Google’s latest spending outlook gave investors another reminder that the AI boom is expensive before it is obviously profitable.

The Verge reports that Google increased its capital-spending estimate to as much as $205 billion, up from a prior projection of as much as $190 billion. Even the low end of the new range is higher than the company had previously signaled.

The concern is not that demand for AI has disappeared. It is that building the data centers, chips, networking, and power capacity behind modern AI services requires enormous upfront investment. Investors are asking how quickly those costs turn into durable revenue.

The practical consequence is a more cautious market mood around AI infrastructure. Big technology companies can afford the buildout, but they still have to justify it to shareholders. The next phase of AI competition may be shaped as much by capital discipline as by model capability.