ChatGPT, Claude, and Grok simultaneously hit by major outages
On September 3, users around the world reported widespread problems with several of the largest AI services—ChatGPT, Claude, and Grok. Official status pages confirmed the outages: issues affected web versions, apps, and certain APIs.
Data from monitoring service Downdetector shows a sharp rise in user complaints. People report errors when sending queries, inability to get responses from models, issues loading services, and increased latency.
Almost simultaneously, OpenAI, Anthropic, and xAI acknowledged their own technical problems.
What’s going on with ChatGPT
OpenAI reported elevated error rates in both ChatGPT and Codex. The company began investigating the incident and later stated it had applied measures to remediate the issue and was monitoring recovery.
According to OpenAI’s official status page, the problems affected 15 ChatGPT components and four Codex components. The company has not yet disclosed the cause of the current outage.
This is not OpenAI’s first technical incident in recent days. Earlier on September 3, the company also recorded elevated error rates in ChatGPT Work mode, but that outage was resolved.
Claude hit by errors across multiple models
Anthropic also officially confirmed serious problems affecting Claude. The first reports of elevated error rates appeared around 13:26 UTC.
Initially, the outage affected Claude Mythos/Fable 5.1, Mythos/Fable 5, Opus 5, Opus 4.8, and Opus 4.6. Problems extended to Claude.ai, the Anthropic API, Claude Code, and Claude Cowork.
The company said it had identified the cause and was working on a fix. Later, Anthropic reported partial recovery: most models returned to normal error levels, but issues persisted for Opus 4.8 and Opus 5.
Grok stopped working normally across several platforms
The developer of Grok, xAI, also confirmed a large-scale outage. According to xAI’s official status page, model issues occurred on the Grok web version, iOS and Android apps, Grok inside social network X, Grok Build, and enterprise plugins.
The outages also affected xAI’s US API infrastructure. The company’s European API continued to operate.
xAI said it was investigating the incident and working to restore service as quickly as possible. For the Grok web version, the start of the outage was recorded at approximately 13:30 UTC.
Cause of simultaneous outages still unknown
At the time of publication, there is no confirmation that the issues with ChatGPT, Claude, and Grok are related. Each company is investigating its own technical incident, and no official common cause for the simultaneous disruption of major AI services has been given.
Therefore, the coincidence in timing alone does not indicate a single infrastructure incident, a cyberattack, or a problem with a common cloud provider.
The situation remains fluid: some functions are gradually being restored, but individual models and platforms may still have errors.
Sources: Downdetector, OpenAI Status, Claude Status, xAI Status
