Outage recap · OpenAI

ChatGPT outage — October 7, 2026

On October 7, 2026, ChatGPT had 5 reported incidents on OpenAI's official status page. Below is what was reported — impact, timing, and the AI services the same upstream outage could affect.

new thread failed on the latest desktop app version

major
Started
Oct 7, 1:31 AM EDT
Resolved
—
Duration
end time not published

What our checks recorded

We checked ChatGPT 24 times in the two hours after OpenAI's start time (it published no end time), once every five minutes. Every one of them reached ChatGPT's site, so this incident did not show up as a failed page load. Our check loads the public site; it cannot see an API error or a broken feature. Median response time was 252 ms, against 266 ms over the 24 hours before. OpenAI's feed showed the incident by Oct 7, 1:35 AM EDT (we read it every five minutes).

Updates are OpenAI's own words from its status page (status: resolved); we never add an inferred cause.

Dot in Codex may fail to create new threads

minor
Started
Oct 7, 12:31 PM EDT
Resolved
—
Duration
end time not published

What our checks recorded

We checked ChatGPT 24 times in the two hours after OpenAI's start time (it published no end time), once every five minutes. Every one of them reached ChatGPT's site, so this incident did not show up as a failed page load. Our check loads the public site; it cannot see an API error or a broken feature. Median response time was 222 ms, against 251 ms over the 24 hours before. OpenAI's feed showed the incident by Oct 7, 12:35 PM EDT (we read it every five minutes).

Updates are OpenAI's own words from its status page (status: resolved); we never add an inferred cause.

Degraded performance in Codex and Work mode: Dot turn failures, delayed responses, and reduced proactivity

minor
Started
Oct 7, 1:21 PM EDT
Resolved
Oct 7, 2:23 PM EDT
Duration
1h 2m

OpenAI's updates (2)

  1. investigating · Oct 7, 1:21 PM EDT

    We are investigating the issue for the listed services.

  2. resolved · Oct 7, 2:23 PM EDT

    All impacted services have now fully recovered.

What our checks recorded

We checked ChatGPT 12 times from OpenAI's start time to its resolution, once every five minutes. Every one of them reached ChatGPT's site, so this incident did not show up as a failed page load. Our check loads the public site; it cannot see an API error or a broken feature. Median response time was 271 ms, against 251 ms over the 24 hours before. OpenAI posted the incident at Oct 7, 1:21 PM EDT.

Updates are OpenAI's own words from its status page (status: resolved); we never add an inferred cause.

Elevated errors with GPT-5.6 Instant in FedRAMP workspaces

minor
Started
Oct 7, 7:42 PM EDT
Resolved
Oct 7, 8:03 PM EDT
Duration
21 min

OpenAI's updates (3)

  1. investigating · Oct 7, 7:42 PM EDT

    We have identified that users are experiencing elevated errors for the impacted services. We are working on implementing a mitigation.

  2. monitoring · Oct 7, 7:48 PM EDT

    We have applied the mitigation and are monitoring the recovery and no longer seeing any errors.

  3. resolved · Oct 7, 8:03 PM EDT

    All impacted services have now fully recovered.

What our checks recorded

We checked ChatGPT 4 times from OpenAI's start time to its resolution, once every five minutes. Every one of them reached ChatGPT's site, so this incident did not show up as a failed page load. Our check loads the public site; it cannot see an API error or a broken feature. Median response time was 162 ms, against 246 ms over the 24 hours before. OpenAI posted the incident at Oct 7, 7:42 PM EDT.

Updates are OpenAI's own words from its status page (status: resolved); we never add an inferred cause.

Some APAC users are seeing errors in ChatGPT Work, conversations, and GPTs.

minor
Started
Oct 7, 11:17 PM EDT
Resolved
Oct 8, 12:51 AM EDT
Duration
1h 34m

OpenAI's updates (2)

  1. investigating · Oct 7, 11:17 PM EDT

    We are investigating the issue for the listed services.

  2. resolved · Oct 8, 12:51 AM EDT

    All impacted services have now fully recovered.

What our checks recorded

We checked ChatGPT 19 times from OpenAI's start time to its resolution, once every five minutes. Every one of them reached ChatGPT's site, so this incident did not show up as a failed page load. Our check loads the public site; it cannot see an API error or a broken feature. Median response time was 223 ms, against 238 ms over the 24 hours before. OpenAI posted the incident at Oct 7, 11:17 PM EDT.

Updates are OpenAI's own words from its status page (status: resolved); we never add an inferred cause.

The AI cascade

ChatGPT depends on OpenAI. When that upstream has an incident, it doesn't hit ChatGPT alone — these AI tools share the dependency and can degrade at the same time:

CursorWindsurfGitHub CopilotTabnineSourcegraph CodyClineRoo CodeContinue

AIDownCheck is not affiliated with OpenAI. This recap summarizes incidents published on OpenAI's official status page.