On September 3, 2026, OpenAI, Anthropic, and Google simultaneously experienced disruptions across their flagship AI services: ChatGPT, Claude, and Gemini. OpenAI bore the most documented impact, with elevated error rates detected at 10:58 AM UTC affecting 15 ChatGPT components and 4 Codex components, resolved within 24 minutes. Anthropic’s Claude Sonnet 5 faced elevated errors for approximately 25 minutes on the previous day, while Gemini endured the longest outage at roughly 2 hours. These simultaneous failures highlight that despite a reported 99.9% uptime, AI providers collectively experienced dozens of similar outages in 2026, totaling nearly 9 hours of annual downtime per service.
Source: Read the original article
