Claude is experiencing an outage, with users encountering elevated errors when sending requests to multiple Anthropic AI models.
The incident began on September 3, 2026, at around 9:41 AM ET, and is affecting several Claude models, including the company’s latest Mythos, Fable, and Opus models.
According to Anthropic’s status page, the company initially said it was investigating elevated errors on requests to Claude Mythos 5.1, Claude Fable 5.1, and Claude Opus 5.
At around 9:41 AM ET, Anthropic said it had identified the cause of the elevated errors and was working on a fix.
A few minutes later, Anthropic published a more complete list of affected models, confirming that the outage impacts Mythos/Fable 5.1, Mythos/Fable 5, Opus 5, Opus 4.8, and Opus 4.6.
For users, the outage can result in requests failing to complete or returning errors when using one of the affected Claude models.
Anthropic said at 10:49 AM ET that it was continuing to work on a fix for the issue, and the outage remains ongoing at the time of writing.
Overall prevention scores can hide what happens after initial access. Once attackers are using valid credentials, prevention drops sharply.
The Blue Report 2026 measures defenses technique by technique across 338 million simulations run in customer production environments.
