Claude Outage Disrupts Thousands as Anthropic Races to Fix Elevated Error Rates Across AI Services

Claude Outage Strikes Thousands of Users Worldwide

Anthropic's Claude AI services experienced a major outage on September 29, 2026, disrupting access for tens of thousands of users across multiple platforms1

1

2

2

. The service disruption began around 10 a.m. ET, with Down Detector reports spiking rapidly as users encountered persistent error messages3

3

. More than 11,000 U.S. users reported issues through Down Detector, with some counts showing over 12,000 reports by 7:34 a.m. PT4

4

. The widespread outage affected users globally, including in India and the United States, impacting both consumer-facing applications and developer infrastructure5

5

.

Anthropic Confirms Elevated Error Rates Across AI Services

Anthropic acknowledged the technical issue within 20 minutes of the initial reports appearing on the Claude Status page1

1

. The company confirmed it was investigating elevated error rates affecting virtually all of its AI services, including Claude.ai desktop and mobile apps, platform.claude.com, the Claude API, Claude Code, and Claude Cowork3

3

. Claude AI, Claude Code, and Claude Cowork were all experiencing partial outages, while Console and API tools showed what Anthropic described as degraded performance1

1

. Notably, Claude for Government remained operational throughout the service disruption1

1

.

Authentication Issues and Error Messages Plague Users

Users attempting to access Anthropic's AI tools encountered multiple types of error messages during the outage. Many received "Request couldn't be processed. You can try again" notifications when trying to chat with Claude on web, iOS, Android, or desktop platforms2

2

. Authentication issues proved particularly problematic, with some users seeing "Claude unable to authenticate" error messages4

4

. The technical issues manifested as failed requests when sending messages or triggering prompts, login hiccups, loading failures where past conversations failed to display, and API timeouts3

3

. Some users found themselves trapped in endless loops when attempting to create new accounts, effectively locked out of the service entirely2

2

.

Anthropic Applies Mitigation as Full Fix Remains Pending

Anthropic engineers began work on addressing the widespread outage immediately after confirmation. The company applied an initial mitigation at 10:40 a.m. ET, which resulted in error rates dropping substantially1

1

. However, Anthropic acknowledged this mitigation wouldn't fully restore AI services to normal operation1

1

. The Claude Status page indicated that "some sign-in attempts, account actions and Claude Code and Cowork sessions are still failing" even after the initial fix1

1

. Anthropic noted the issues were intermittent, meaning retrying requests might occasionally succeed, though overall reliability remained severely degraded3

3

. The company provided no specific timeline for when complete resolution would occur4

4

.

Impact on Developers and AI Infrastructure Reliability

The Claude outage extended beyond casual users to significantly impact developers relying on Anthropic's AI infrastructure. Both everyday users and developers faced persistent error messages and broken API calls as the service disruption spread from Claude.ai to platform services3

3

. For development teams whose technology stacks depend on the Claude API, the outage meant work delays until Anthropic's engineering team could roll out a complete patch3

3

. This incident marks another challenge for AI infrastructure reliability, following a major outage earlier in September that caused various AI providers, including Anthropic's Claude tools, to struggle maintaining live services1

1

. While competing services ChatGPT and Google Gemini saw small spikes in outage reports on Down Detector during the same timeframe, their issues affected only hundreds of users compared to the tens of thousands reporting problems with Claude1

1

.