On September 29, 2026, Claude had a significant disruption starting at 14:23 UTC. It lasted about two hours and was resolved by 16:27 UTC. Anthropic has since marked it as resolved. That is the full picture the official source gives us, and we will not dress it up with detail that was not published.
Who this kind of outage hits
If you use Claude inside an n8n workflow, a customer-facing tool, or any automated process, you probably did not get a warning. You were just waiting. The work stopped moving and nothing told you why. Most operators find out the same way: a customer writes in to say something is broken, or they check their own tool hours later and notice a backlog of failed runs. By then the outage is already over, but the damage, the missed responses, the stalled automations, the confused customers, is already done.
That gap between “the tool stopped” and “I found out” is the real problem. It is not the outage itself. It is the silence around it.
Why this is especially rough without an engineering team
If you have engineers, someone is probably watching logs. But most people building on Claude are not running an engineering team. There is no error on the screen. The workflow just sits there. A customer-facing feature quietly returns nothing. You find out when someone complains, or when you happen to open the tool and notice the queue. There is no alert, no plain-language explanation, just a gap in the work where output should have been.
That is the shape of a quiet outage. It is not dramatic. It is just invisible until it is not.
Timeline
- 14:23 UTC - Claude started having a bigger problem.
- ~16:27 UTC - Claude recovered, about two hours later.
- Anthropic has confirmed the issue is resolved.
How a watcher catches this before your users do
Anthropic, like most tools, publishes a public status page. When something goes wrong, they update it. The gap is that most people are not watching that page. They are working.
NoCrash reads Claude’s public status page every minute. The moment Anthropic marks something as having trouble, NoCrash translates that into plain English and sends you a heads-up. Not a raw status code. Not a technical summary. A calm sentence that tells you what is going on, in the same place you watch everything else you build on.
It also watches the things you ship. If you have n8n workflows, NoCrash watches those too. If you give it a URL or add a small JS snippet to your app, it watches that as well. So if the problem is on your side rather than Claude’s, that surfaces too.
What NoCrash does not do: it does not find the outage before Anthropic’s own status page does. It reads that page and tells you quickly, in words you can act on. That is the whole thing. It turns “my customer told me two hours later” into “I got a plain heads-up within a minute of Anthropic’s own report.”
For the authoritative account of this outage, see Anthropic’s own status page: https://stspg.io/br61xzj05pp5