Former CIA Advisor Warns AI Infrastructure Race May Create Conditions for Its Own Collapse

A former CIA senior advisor warned that massive AI infrastructure buildout may create conditions for financial disruption. Hundreds of billions poured into data centers while actual AI application revenue lags far behind.

Former CIA Advisor: AI Infrastructure Race May Create Its Own Collapse

A former CIA senior advisor published a stark analysis on March 29, arguing that AI infrastructure's frantic buildout may be creating conditions for financial disruption. Core thesis: AI investment is driven by FOMO rather than rational ROI calculations.

2025 global AI capex exceeded $300B with US hyperscalers alone spending $200B+. However, actual AI application revenue after infrastructure costs is estimated at only 15-20% of investment. The advisor draws parallels to the 1990s fiber optic buildout that led to massive overcapacity and bankruptcies including WorldCom.

Counter-argument: unlike fiber overcapacity, AI compute demand is growing exponentially (China's 140T daily tokens evidence). The question isn't whether infrastructure is needed, but whether the pace is sustainable.