The hyperscaler debate has changed. When we first examined the economics of the AI infrastructure buildout in our original Hyperscalers: Capex vs Returns analysis, we argued that the sector was moving from a capital-light cloud model toward a more capital-intensive era, with investor attention increasingly shifting from growth to free cash flow and returns. Six months on, AI demand is no longer the biggest question. The bigger issue is what it costs to meet that demand, and when unprecedented infrastructure investment begins translating into stronger cash returns for shareholders. Capex continues to accelerate faster than revenue, while the expected free cash flow recovery has moved further out, putting even greater focus on the economics and ultimately the returns behind the AI buildout.

General
Value-Realization Framework for Turning Enterprise AI into Measurable Performance
Many companies have deployed AI, but few can show that it changed how the business performs. Gururaj Bhat outlines a six-part framework for turning AI investments into measurable results.


