Anthropic’s Claude AI had a service disruption on September 22, 2026, with users in several regions unable to access the popular artificial intelligence platform. The issue also seems to affect Claude Code, Anthropic’s AI-based coding tool, and developers who rely on it for programming and other technical work.
The latest disruption comes amid a series of service issues on Anthropic platforms in the last few weeks. More than 1,600 reports were recorded during the recent outage, according to outage-tracking data. Users reported that Claude was either unavailable or not functioning normally, and some experienced problems with Claude Code.
The September 22 incident was not immediately accompanied by a detailed technical explanation from Anthropic. At the time of reporting, the company had not publicly confirmed the exact cause of the latest disruption or provided a detailed breakdown of which Claude models were affected. This means that reports from users and outage-monitoring platforms provide much of the available information about the scope of the incident.
Claude AI Outage Leaves Users Facing Access Problems
The disruption affected users who attempted to access Claude on different services. Some users reported problems with the main Claude platform and others had trouble using Claude Code. The problems were reported in different regions, which shows that the problem was not limited to one local network or geographic location.
Outage monitoring platform Downdetector received more than 1,600 reports during the disruption. Such reports can change quickly as users are affected and later they return to the service, so they may not represent the total number of people affected.
For users who rely on Claude for everyday AI assistance, the disruption meant that conversations, prompts, and other tasks may be temporarily interrupted. For developers, it might be more important, as Claude Code is currently part of software development and debugging tasks.
Latest Claude Status: Services Restored
The good news for Claude users is that Anthropic’s system status page later indicated that its services had returned to normal operation. An independent monitoring service also listed Claude as operational when checked on September 22.
But individual users may still experience problems even after the server-side outage was resolved. Temporary connection issues, browser issues, account-related errors, or regional network disruptions can persist after the main server-side incident has ended.
Users who still encounter problems should check Anthropic’s official system status information before spending a lot of time troubleshooting their own devices or internet connections.
September 3 Claude Outage Had A Wider Model Impact
The September 22 disruption follows another major Claude outage on September 3, 2026. That earlier incident is more important since several Claude models and services were affected.
Anthropic's status updates during the September 3 incident indicated that Claude Mythos 5.1, Claude Fable 5.1, and Claude Opus 5 were among the models affected. The list was expanded in subsequent updates to Mythos and Fable 5.1, Mythos and Fable 5, Opus 5, Opus 4.8, and Opus 4.6.
Both claude.ai and the Claude API were reported as experiencing partial outages during that incident, while the Claude Console was working. Anthropic said in the past that it had found the cause and was working on a solution, but details of the underlying problem were not made public.
Claude for Government was reported as unaffected by that incident.
Recent Claude Service Issues
The latest disruption also comes after several shorter incidents recorded in Anthropic's system status history. Service problems with specific Claude models and features were reported on August 24, September 10, September 11, and September 15.
The repeated incidents highlight how dependent users and businesses have become on large-scale AI services. Even short interruptions can affect people using AI for writing, coding, research, customer support, and automated workflows.
Anthropic also recently began its latest generation of Claude models, with Claude Fable 5.1 and Claude Mythos 5.1 going on sale September 1. The company put the models to work with coding and knowledge-work tasks, making service reliability very important for professional and developer users.
What Should Claude Users Do If The Service Is Down?
If Claude stops responding, users can first check if this problem is widespread or limited to their own connection. Reloading the webpage or restarting the application may resolve temporary client-side issues. Logging out and signing back in may also help to identify account-session problems.
Users can also try another internet connection, such as switching from Wi-Fi to mobile data. Checking whether Claude.ai, Claude Code, or the API has been affected might also provide useful information in case an incident occurs but one service does not take down the whole Anthropic platform.
The first step in a suspected widespread outage is to check Anthropic's system status page. If the company is already investigating an incident, refreshing the page several times, changing some settings, and reinstalling applications will not help a server-side issue.
For developers, this distinction is especially important. API failures can appear to be caused by local code, authentication settings, or application configurations when the actual problem is on the provider’s side.
Why Claude Code And API Outages Matter
A disruption of Claude Code can have a bigger impact on developers than a regular chatbot outage. Developers can use AI coding tools to generate or modify code, investigate bugs, review projects, and troubleshoot development issues. If the service is unavailable in the middle of a workflow, ongoing tasks can be delayed.
API customers may also go through another layer of potential disruption. Claude can be used for customer-facing applications, internal tools, automated workflows, and other software systems. If API requests fail during an outage, other dependent services may also be affected.
For now, Anthropic’s status indicates that Claude services are functioning normally following the September 22 disruption. But this incident is just one of a number of service interruptions in the past few weeks, which makes it more important to keep track of provider status pages when AI tools are part of critical digital workflows.
