OpenAI Outage History

From early on 3:45 AM UTC, OpenAI users reported chatgpt.com access issues, with ChatGPT login and related services showing degraded behavior. Users experienced timeouts, failures to load conversations, and errors when uploading or processing files. The disruption touched a broad set of OpenAI components, including ChatGPT / Login and a wide range of APIs such as Batch, Files, Sora, Moderations, Realtime, Audio, Embeddings, Images, Chat Completions, Responses, and Fine-tuning. The incident appeared in waves across different services and regions, producing a global sense of disruption rather than a single regional outage.

Impact began around Jun 2, 2026 11:15 PM UTC as OpenAI's ChatGPT Pro users encountered elevated errors, including Thinking Failed messages when using the Pro model. The disruption affected a broad set of components including ChatGPT Conversations, Login, File uploads, Voice mode, GPTs, Search, Deep Research, Image Generation, Agent, ChatGPT Atlas, Connectors/Apps, and the Compliance API, leading to degraded performance and intermittent failures for Pro users. The issue appeared global, with reports from multiple regions around the world.

OpenAI's ChatGPT conversations were experiencing elevated errors for guest users, disrupting the ability to start or continue chats. The issue appears to have begun around 12:06 PM UTC and coincided with a noticeable uptick in guest-user reports starting near 12:00 PM UTC. Some users reported failures when sending messages, resulting in a degraded conversation experience.

OpenAI's status updates indicate that they identified the issue and implemented mitigation around 12:51 PM UTC, with monitoring ongoing. The ChatGPT / Conversations component returned to normal by 1:16 PM UTC, signaling the incident was resolved after about one hour. The episode affected guest accounts and was characterized by elevated error rates rather than a full outage, as indicated by the status updates and user reports.

OpenAI experienced a brief degradation in ChatGPT availability for Free users, starting around 4:26 PM UTC due to capacity constraints on ChatGPT / Conversations. Free-tier access was reduced and some users encountered slower responses or intermittent errors during the impact window. The degradation persisted for about an hour before service began returning to normal.

Status updates indicated a gradual recovery. By 5:01 PM UTC the page noted that ChatGPT success rates were back up to normal functionality, and the ChatGPT / Conversations component moved back to full availability by 5:32 PM UTC. Overall, users on the free tier experienced degraded access before full restoration.

In terms of impact, the disruption affected free-tier access to ChatGPT, with user reports describing interruptions and unresponsiveness during the window. The issue did not indicate a broader regional pattern and was resolved within the hour.

OpenAI acknowledged a brief degradation affecting a broad range of ChatGPT features, with users reporting login problems and intermittent access. The impacted components included Conversations, Login, Compliance API, Search, File uploads, Voice mode, GPTs, Image Generation, Deep Research, Agent, ChatGPT Atlas, and Connectors/Apps. The incident began at 1:22 PM UTC and OpenAI stated they were investigating the issue.

All affected services were restored by 1:31 PM UTC. The disruption lasted about nine minutes, after which normal operation resumed. No further public root-cause details were provided in the status updates.

StatusGator detected issues at May 29, 2026 3:05 AM UTC, about 11 minutes before OpenAI acknowledged the problem at May 29, 2026 3:16 AM UTC. OpenAI's ChatGPT experienced an outage affecting login and conversation features; users reported being unable to log in or create accounts, prompts not loading, and conversations not loading or behaving inconsistently. The issue extended to prompts and image loading in some cases, with a broad set of users reporting across many regions. OpenAI issued status updates indicating problems with ChatGPT / Login and APIs / Login, followed by ChatGPT / Conversations as the incident evolved, signaling a partial outage rather than a total blackout. Mitigation steps were applied around 3:32 AM UTC with monitoring continuing as recovery progressed.

By 5:01 AM UTC, recovery was underway and services returned to normal for most users; OpenAI confirmed the affected components were stabilized and logging and conversation experiences were functioning normally again. The incident generated a high volume of user reports in the early phase, indicating broad impact across regions, and the service was back to normal within about two hours of the initial impact.

OpenAI users attempting to subscribe to the Business plan on the web or mobile web faced checkout problems, with some transactions failing or processes stalling. The disruption began at 8:47 PM UTC and persisted for several hours, with affected activity centered on the Business plan checkout flow. User reports originated from North America, including Vancouver and Maple Ridge in Canada and Merced, California, as well as Mooresville, North Carolina, suggesting a regional pattern rather than a global outage. The issue impacted signups and renewals for the Business plan during the window of disruption.

OpenAI experienced a broad outage affecting FedRAMP logins and API performance, beginning at 5:01 PM UTC when FedRAMP users who had recently logged out encountered login issues. The initial impact covered the FedRAMP / FedRAMP and ChatGPT / Login components, and soon expanded to elevated latency and errors across the API surface, including Audio, Embeddings, and Chat Completions, leading to slower responses and intermittent failures for users around the world. Reports came from multiple regions, indicating a global disruption rather than a localized fault.

Users began experiencing degraded access to OpenAI Codex starting 4:42 PM UTC as an uptick in Codex rate limits affected multiple components, including Codex App, Codex API, Codex CLI, Codex VS Code extension, and Codex Web. Reports described failures to connect, slow responses, or content loading errors across Codex-powered tools, with incidents appearing in IDEs, the web interface, and API clients.

OpenAI introduced a mitigation at 7:16 PM UTC and monitored recovery as traffic normalized. By 11:01 AM UTC, Codex services had returned to normal operation. The event was a rate-limits-related disruption rather than a total outage, with gradual improvement as usage patterns stabilized.

Users began reporting degraded performance for ChatGPT 5.5 Thinking, with elevated latency and intermittent errors affecting the ChatGPT / Conversations component. Some interactions timed out and responses were slower than usual. The issues were first seen around 9:15 PM UTC, and OpenAI acknowledged a degraded state around 10:06 PM UTC as the status page updated. OpenAI implemented mitigations around 10:26 PM UTC and continued monitoring; the ChatGPT / Conversations component returned to normal by 11:26 PM UTC. The degradation persisted for roughly an hour to an hour and a half, with the impact mainly on chat interactions.