Anthropic’s AI assistant hit a wave of errors before a quick fix restored service
For a stretch this afternoon, one of the most popular AI assistants left a lot of people staring at error messages.
Anthropic’s Claude AI hit an outage today with users flooding outage-tracking sites to report that the service had stopped working. The disruption was brief, but widespread enough to spark frustration, jokes and a few conspiracy theories before the company stepped in with a fix.
What happened
The trouble started just after 1 p.m. ET, according to reports on Downdetector. Complaints climbed quickly, peaking at more than 2,100 around 1:26 p.m. before tapering off. About half of the issues were tied to Claude Code, the company’s coding tool, while others involved the main Claude chat interface and trouble logging in.
Anthropic acknowledged the problem on its status page shortly before 1:30 p.m. ET, saying it was investigating elevated errors affecting many of its models across the Claude web interface, its API, Claude Code and Claude Cowork. In other words, the hiccup touched a broad swath of the company’s products at once.
A fast fix
The good news for users is that the outage did not last long. By around 2 p.m. ET, Anthropic said it had implemented a fix and was monitoring the results. The flood of complaints receded almost as quickly as it arrived, with Downdetector reports falling from their peak to roughly 500 within about 15 minutes. By later in the afternoon, independent monitoring sites listed Claude as operating normally again.
Outages like this are not unusual for major AI platforms, which juggle enormous demand and complex infrastructure. Claude incidents typically resolve within about 94 minutes, and this one fit that pattern.
A rumor that wasn’t true
The downtime also gave users on social media and outage forums an excuse to speculate. Some joked that the errors meant Claude Fable 5, a recently released model, was making a comeback. That, however, was not the case.
Fable 5, described as the mass-market version of Anthropic’s more powerful Mythos 5 model, debuted only last week. Anthropic then pulled access to it on Friday after the U.S. government imposed export controls that barred foreign nationals from using the model. The Trump administration pointed to what it described as a jailbreak that could let users get around the model’s safeguards. Anthropic has pushed back, saying the issue is not as serious as the government has claimed and objecting to the order.
None of that, though, had anything to do with today’s errors, which the company framed as a routine technical issue rather than anything tied to a specific model or policy dispute.
For now, Claude appears to be back to normal, and its many users can return to their chatbots, coding assistants and everything in between.
Source: AOL