Mass Claude Outage: Anthropic Confirms Problems Across Several Services

Anthropic has confirmed a major outage affecting Claude. Users complained that they couldn't log in and that responsiveness had worsened across several of the company's services at once.

According to the status page, the incident began on August 16, 2026 at around 21:58 UTC and affected Claude.ai, Claude Code and Claude Cowork. Anthropic first said it was investigating an issue that prevented some users from authenticating with these services. A few minutes later, at 22:07 UTC, the company added that it was observing degraded performance on Claude.ai and platform.claude.com.

In practice, the outage translated into an inability to sign in, Claude hanging while loading, unfinished requests and other errors when trying to reach the affected services. At the time of publication, Claude.ai, Claude Code and Claude Cowork were listed on the status page as experiencing a “major outage,” while Claude Console and the Claude API remained operational.

Anthropic did not disclose the cause of the trouble — both incidents were still under investigation as this note was being written. The company limited itself to brief updates on the progress of its inquiry, naming neither a recovery timeline nor the source of the problem.

Outages like this are nothing new for Anthropic: the service has gone offline worldwide before. This time readers greeted the news with irony — in the comments on the original post, someone couldn't resist a “Terminator” reference: Skynet had become self-aware, they joked, and Judgment Day had arrived. Reality, however, turned out to be far more mundane — an ordinary technical glitch, albeit one that hit several of the company's key products at once.

Source: Bleeping Computer