OpenAI said it was monitoring a gradual recovery after a worldwide ChatGPT outage on July 19 that disrupted access for users across multiple regions. The incident affected people in the United States, Europe, Ukraine, India, Japan, Australia, and other locations, pointing to a broad service problem rather than a local issue.
During the disruption, users reported several common symptoms. Some saw 502 Bad Gateway and 503 Service Unavailable messages, while others experienced slow responses or an unresponsive chat interface. Those signs suggested that both availability and performance were impacted during the outage window.
According to the available details, OpenAI deployed a mitigation roughly two hours after the incident began. The company then said it was watching a gradual recovery, but it did not share the root cause of the outage or give a full timeline for complete restoration.
The event highlights how widely ChatGPT is used around the world and how quickly outages become visible across regions. While service appeared to be improving, OpenAI had not yet provided a detailed explanation for what triggered the worldwide disruption.
