Anthropic investigates sign-in errors on 4 Claude services

Anthropic reports that it investigates more errors than usual when users sign in to Claude.ai and Claude Code. The same incident affects the Claude Console and Claude Cowork, 2 other services of Anthropic. The short notice has no numbers, no cause and no start time, and Anthropic tells only that it will give an update later.

What is affected

The notice shows 4 services:

• Claude.ai
• Claude Console
• Claude Code
• Claude Cowork

The proof

The notice has no numbers and no test results. Its status is Investigating. Anthropic gave no time for the start of the errors.

What is not known

The notice does not give the cause. It does not tell how many users have errors, or when the repair will come. It does not tell if data or accounts are at risk.

Insight: One sign-in error can stop 4 Claude tools

The notice names 4 services with one sign-in problem. In October 2026, Anthropic added agent workflows with a maximum of 1,000 agents. It also added a tool that checks code at each change. This reads as Claude becoming part of the work of many teams, and one error can stop more of it. The notice does not connect this incident to the bad Claude Code add-on of October 2026.