Claude Down Twice on Friday 13 November as Logins Fail

Anthropic's AI assistant Claude went down twice on Friday 13 November 2026, in a bad day for the service that first locked users out of their accounts and then disrupted the tools developers and businesses build on.
The first failure began at 09:58 UTC, when Anthropic logged a major incident on its status page and said it was investigating a problem with users authenticating to Claude.ai and Claude Code. The fault affected four services: the Claude.ai chatbot on the web and in its apps, the Claude Code coding agent, the Claude Cowork office agent, and the Claude Console where developers manage accounts and API keys.
A fix was deployed at 11:05 UTC, about 67 minutes after the first report, and Anthropic marked the incident resolved at 11:58 UTC, two hours after it began. The company has not said what caused the failure, how many people were affected, or whether any work was lost.
Second outage hits the Claude API
Just over an hour later, a second and wider incident began at 13:09 UTC. This time the problems reached the paid Claude API itself, along with Claude Code, connectors, Claude Security and Claude Managed Agents. Anthropic reported recovery on Claude.ai and the API at 14:36 UTC, while several services stayed degraded, and marked the incident resolved at 15:38 UTC, about two and a half hours after it opened.
The afternoon failure had sharper consequences than the morning lockout. Anthropic warned during the incident that scheduled tasks and routines could run more than two hours late and that some runs had been skipped, and adding a new connector was temporarily unavailable. A login failure blocks people at the door of the chat app; an API failure means work that companies rely on does not run, or runs hours late.
Not linked to the Microsoft 365 outage
Some coverage tied Claude's problems to the Microsoft 365 outage that began on Thursday 12 November and disrupted Outlook. Anthropic has not made that link, and nothing on its status page supports it. The company named Microsoft in only one incident: a separate failure of its Microsoft 365 connector, running since 17:52 UTC on Thursday, which Anthropic said was caused by a Microsoft 365 service incident. That connector lets Claude read mail, calendars, SharePoint and Teams chats, and its mail, calendar and search tools were still failing for most requests even after sign-in was restored.
The timing of the morning outage hit Europe's working day hardest. It began at 9:58am in London and 10:58am in Berlin, and at 8:58pm in Sydney, while it was 4:58am in New York. Anthropic's status page is the fastest place to check whether a problem is yours or theirs: if an incident is open, signing out to try to fix an error can leave you locked out until it ends, as the company's own advice during an earlier outage in September made clear.
Friday's failures were the second major sign-in disruption in seven weeks, after a one-hour outage on 29 September in which Anthropic warned that some messages may not have been saved. With Anthropic about to price its stock market listing, customers and investors will want to know what broke. So far, the company has published no post-incident report for either Friday outage and has given no cause for either one. Sources: MadRobot and Agenccy, both reporting from Anthropic's public status page at status.claude.com.