September 3 began like another Thursday, till it didn’t. Inside roughly 90 minutes, ChatGPT, Claude, Grok, and even Microsoft’s personal Copilot have been degraded or darkish, knocked sideways by a failure in Microsoft Azure’s East US area. Downdetector logged greater than 37,000 stories for ChatGPT alone, greater than 1,300 for Claude, and roughly 1,365 for Grok. OpenAI’s standing web page flagged elevated errors throughout 15 ChatGPT parts and 4 Codex parts, and by the point engineers had mitigations in place, mixed report counts had climbed previous 66,000.
What made this occasion outstanding was not the size of any single outage, however its simultaneity. Three aggressively competing AI labs—OpenAI, Anthropic, and xAI—every spend billions differentiating their fashions, but all three buckled at practically the identical second as a result of they shared the identical regional dependency. Gemini, notably, stayed largely upright as a result of Google runs its flagship assistant by itself vertically built-in cloud. The outage that took down its rivals had no assault floor inside Google’s stack. That’s not luck; that’s structure.
The lesson buried within the particulars of that morning is that this: A single cloud area turned unhealthy, and 4 of probably the most outstanding AI companies on the planet, owned by 4 completely different corporations, fell over collectively. Copilot’s involvement is probably probably the most telling information level. Microsoft’s personal first-party assistant runs on Microsoft’s personal cloud, and it nonetheless had a tough morning. When the home itself shakes, you realize the failure sits deep in shared infrastructure layers, not in any tenant’s configuration.
