Unprecedented Simultaneous Outages Hit Global AI Services... ChatGPT, Claude, and Grok All Go Down

Major AI Platforms Experience Outages on the Same Day
Over 37,000 Outage Reports for ChatGPT Alone

Major artificial intelligence (AI) services around the world, including ChatGPT, Claude, and Grok, experienced simultaneous access disruptions within a similar timeframe. Not only AI chatbots but also coding tools were affected, causing inconvenience for users around the globe for a period of time.

According to Yonhap News on September 4, citing Forbes, The Verge, and Axios, consecutive access errors occurred on key generative AI services such as OpenAI's ChatGPT, Anthropic's Claude, and xAI's Grok.

ChatGPT users reportedly experienced errors not only in conversation generation but also in various functions such as login, file upload, search, voice mode, and image generation. OpenAI also confirmed the increased error rates in ChatGPT and Codex through its official status page and took measures to address the issue. EPA Yonhap News

On Downdetector, an online service outage tracking site, reports of disruptions to major AI services surged after 9:00 AM Eastern Time on the 3rd. In particular, reports regarding ChatGPT surpassed 37,000 around 11:00 AM, representing the sharpest increase. During the same timeframe, Claude saw over 1,300 reports, and Grok also received more than 1,300 incident reports.

ChatGPT users reportedly experienced errors not only in conversation generation but also in various functions such as login, file upload, search, voice mode, and image generation. OpenAI also confirmed the increased error rates in ChatGPT and Codex through its official status page and took measures to address the issue.

Major AI services such as ChatGPT, Claude, and Grok, which are competitors, experiencing issues simultaneously within a similar timeframe is considered unusual. Downdetector

After applying mitigation measures, OpenAI monitored the situation as the service recovered, and eventually announced that "the issue that caused increased errors in ChatGPT and Codex has been resolved." The company also noted that some remote Codex users might need to reconnect their mobile devices following the disruption.

The same day, Anthropic’s Claude also experienced errors across several models. Anthropic began investigating issues with some Claude models from 9:00 AM, and after sequentially restoring affected models, announced at 12:16 PM that all reported issues had been resolved. Service interruptions were observed in Claude, Claude Code, and the API, with an Anthropic spokesperson attributing the partial incidents to "infrastructure issues."

xAI's Grok also experienced issues on web and mobile apps. According to The Verge, errors began appearing on Grok around 9:30 AM, and when users submitted questions, a message indicating server overload was displayed. Reuters Yonhap News

xAI’s Grok also encountered disruptions on web and mobile applications. According to The Verge, errors began appearing on Grok around 9:30 AM, with a server overload message sometimes displayed when users submitted questions. Later, xAI recovered the service and linked the disruption to an issue at its Memphis data center.

The impact of the disruptions was not limited to AI chatbots. On the same day, Cursor, an AI coding tool acquired by SpaceX last month, also suffered degraded performance and experienced errors related to OpenAI and Anthropic models. Cursor provides an environment where users can work with various AI models, including those from OpenAI, Anthropic, and Grok.

Google’s Gemini also saw some user reports of disruptions during a similar period. Around 11:00 AM, there were approximately 500 error-related reports regarding Gemini. However, Google did not issue a separate official notice at that time confirming any major service disruptions, and the volume of reports was comparatively smaller than that of other AI services.

It is considered unusual for multiple major, competing AI services—such as ChatGPT, Claude, and Grok—to simultaneously encounter issues within the same time period. Some observers speculated that the disruptions may have been influenced by an outage in Microsoft (MS) Azure's cloud service. At the time, Azure also saw a spike in incident reports, and since major AI companies rely on Azure and other external cloud infrastructure, the possibility arose that a common infrastructure problem was the root cause.

However, as of now, OpenAI, Anthropic, and xAI have not officially identified a common cloud issue as the source of these disruptions. On the contrary, xAI specifically explained the Grok outage as being related to a problem at its own Memphis data center. The Verge also reported that it has not been confirmed whether the outages at the three companies were interconnected.

The disruptions began to be quickly restored around noon, and afterward, most major services—including ChatGPT, Claude, and Grok—returned to normal operations.

This incident is seen as a clear example of how, as reliance on generative AI rapidly deepens in areas such as work, search, and programming, the stability of key AI services—and the underlying data center and cloud infrastructure—has become a critical new challenge.

This content was produced with the assistance of AI translation services.