It was a bit of a test of reliability for Anthropic when, on June 2nd, Claude users in all sorts of places started running into trouble. You had hundreds of them putting up flags about not being able to reach the bot or the site, and you could see the numbers jump on Downdetector. The company has put out a word that they know about it and their engineers are at work on a solution – the kind of thing that shows what’s at stake in the AI assistant game.
Immediate impact and stakes
When an AI tool of this size goes down, it’s more than a minor nuisance. You risk losing developers, you make enterprise clients nervous, and you let your competition in the door. We’ve got confirmation from Anthropic of an incident with a number of models, while on the ground, users were having a hard time even getting a response from any of the Claude services.
This is on top of some other issues with Opus, Sonnet and Haiku of late, so there’s a story here that Anthropic has to put to rest. With ChatGPT and Gemini in the mix, you can’t afford to be anything but steady.
What users saw during the disruption
Most of the noise came from the Claude Chat side of things, so the end-user was where it was felt. Then you had the app, the login process, the API – all of it was in the complaints. If you look at the Downdetector chart, there were a few blips earlier in the day before it really took off around noon.
Here is what we are seeing from the reports:
– 62% had a problem with Claude Chat
– 17% with the app
– 15% couldn’t get to the website
– By 12:41 PM, we were over 230 reports
– Before that, you were looking at one or two at most
Downdetector is only as good as the people who use it, but with Anthropic chipping in to confirm it, you know the disruption was for real and not just in one corner of the world.
Scope and status of the incident
If you check the status page, you’ll see an entry for ‘Elevated errors across multiple models’ that was put up at 06:04 UTC on the 2nd. A little later, at 06:39, it was marked as Identified; the cause was in and a fix was in the works.
The dashboard had a few services under a Partial Outage. They are talking about the Claude Console and the API, and a later note made mention of Claude.ai and Claude Code. As of now, there’s no word on when everything will be back to normal.
Some of the key points from Anthropic:
– Incident was logged at 06:04 UTC, June 2
– Put in the Identified column by 06:39 UTC
– They say the fix is being put in place
– A handful of services are in a partial outage
Market implications and next steps
Claude is the backbone for a lot of consumer, developer and enterprise work, so being up and running is part of the pitch. Even a short spell of unavailability can throw a wrench in the works for someone trying to integrate a new tool, particularly if they are on a clock.
The bottom line is whether Anthropic can win back trust and not let this happen again. Being open about what went wrong and how you fixed it is the only way to head off any second-guessing from customers.
Competitive context
You have to measure up to ChatGPT and Gemini in a market where buyers want to know their tool won’t let them down. Every time an organisation puts AI to work in content or code, an outage is going to be noticed. People are going to be watching to see if the stability holds up before they put all their eggs in one basket.
