The chatbot is said to be experiencing a partial outage which started around 6am this morning, with hundreds of users reporting the issue.

Reports on DownDetector also spiked yesterday (7 April) at around 3:44, with 558 reporting issues with the chatbot.

Anthropic said on the Claude status page that users were seeing an “elevated rate of errors” when using Sonnet 4.6, the model that powers Claude, yesterday.

Other parts of Anthropic’s systems are also suffering from technical issues including Claude API and Claude Code, the organisation has reported.

It appears that while users are still able to access Claude, the system is getting stuck ‘thinking’ and unable to respond to queries. 

Anthropic’s status page noted that even on 6 April users reported elevated errors on Claude.ai. While prior incidents were reported as resolved, it remains uncertain if the current issue is connected.

Most recently, Anthropic said that “The issue has been identified and a fix is being implemented” on its status page.

They are continuing to investigate the issue. 

Recommended reading

• Claude Source Code Leak Reveals Anthropic’s Secret Plans
• Staffers Resist Change as Anthropic Sues US Gov Over AI Contract Quagmire
• OpenAI Backtracks on “Opportunistic and Sloppy” US Government Deal

While an outage may be an unwanted and yet more typical occurrence, Anthropic has continued to deal with political and security battles.

It is still enthralled in a legal spat with the US government over its defence contract and subsequent supply-chain risk designation, and its chatbot’s own source code was recently leaked.