The energy needed to power the AI revolution
AI-summarised brief · reviewed before publication
Millions of workers worldwide experienced a disruption when AI assistant Claude went offline on March 2, 2026, highlighting the fragility of the infrastructure behind AI. The energy needed to power AI is growing exponentially, with global data center electricity consumption projected to nearly double by 2030. This strain is driving hundreds of billions of dollars in grid spending, with homes also requiring reliable, always-on, clean power to support AI-powered functionality.