Four major AI models suffer rare overlapping downtime

On Thursday morning, users of major AI platforms experienced a rare, synchronized service disruption affecting OpenAI, Anthropic, xAI, and Google. The outages, which lasted several hours, caused significant performance degradation across various models. Anthropic reported elevated error rates for its Claude 5 series, including Mythos, Fable, and Opus, starting at 9:23 am ET. The company deployed a fix by early afternoon. Simultaneously, OpenAI faced issues with ChatGPT and Codex, reporting degraded performance beginning at 10:43 am. While mitigation efforts were successful by 12:55 pm, the overlapping nature of these failures across multiple industry leaders has raised questions regarding the underlying infrastructure dependencies of modern generative AI services. Service providers have since restored full functionality, though the root causes for such a widespread, simultaneous failure remain a subject of industry scrutiny.