ChatGPT went down for many users on the morning of July 12, 2026, with iOS users hitting 403 errors when trying to load or continue conversations. The disruption began around 5:29 AM ET, OpenAI confirmed active issues on its status page, and the service was reported operational again by 9:46 AM UTC — a short but visible outage during GPT-5.6 launch week.
ChatGPT is down for a portion of users right now because of a 403 error affecting iOS conversations, first reported around 5:29 AM ET on July 12, 2026. OpenAI’s status page confirmed it was investigating, and monitoring service StatusGator marked ChatGPT operational again by 9:46 AM UTC. The interruption was brief and appears resolved.
- What broke: iOS users hit 403 errors loading or continuing ChatGPT conversations.
- When: Reports surged around 5:29 AM ET on July 12, 2026.
- Status: OpenAI’s official status page confirmed it was investigating the issue.
- Resolved: StatusGator marked the service operational by 9:46 AM UTC.
- Uptime: OpenAI’s status page shows ChatGPT at 99.86% aggregate uptime for April–July 2026.
What the ChatGPT 403 outage actually means
A 403 error means “forbidden” — the server received the request but refused to fulfill it. For ChatGPT iOS users on July 12, that translated into conversations that would not load or continue, rather than the app failing to open entirely. According to OpenAI’s status page, the company acknowledged that “some users may encounter 403 errors when loading or using ChatGPT conversations on iOS” and marked the incident as under investigation.
This was a partial outage, not a full blackout. Many users were unaffected, and the window was short: monitoring service StatusGator recorded ChatGPT as operational again at 9:46 AM UTC. Even so, “ChatGPT down” trended in the UK at roughly 500 searches, showing the interruption hit enough people to spill into social search during peak morning hours.
A separate FedRAMP issue is still open
The consumer iOS outage should not be confused with a longer-running enterprise problem. OpenAI’s status page describes an ongoing FedRAMP workspace incident that has run for about a week, separate from the July 12 iOS errors. In its own words: “Core functionality has been restored but we are aware of known and ongoing issues in FedRAMP workspaces with Codex, workspace analytics, conversation search, searching for custom GPTs, ChatGPT user invites, and the Compliance Logs Platform download endpoint.”
That distinction matters. The iOS 403 errors were a brief consumer hiccup that resolved within hours, while the FedRAMP issues affect government-grade enterprise workspaces and have persisted far longer. Both were visible on the same status dashboard on July 12, but they are different incidents.
What this means for users and for OpenAI
For everyday users, the practical takeaway is simple: a 403 outage like this typically clears on its own, and checking OpenAI’s official status page is faster than guessing. The July 12 incident resolved without user action needed.
The bigger picture is about timing. The outage landed during GPT-5.6 launch week — GPT-5.6 launched last week with the Sol, Terra, and Luna models — a period when traffic and infrastructure load run high. OpenAI’s status page reports 99.86% uptime for ChatGPT across April to July, which sounds strong but works out to roughly ten minutes of downtime per day across a platform used by hundreds of millions. For a company reportedly preparing an IPO roadshow, reliability headlines carry extra weight. Reliability has quietly become a competitive battleground, and readers weighing platforms often factor it into a ChatGPT vs Claude reliability comparison.
What we don’t know yet
OpenAI has not published a root-cause breakdown for the July 12 iOS 403 errors, and no official count of affected users was given. It is not confirmed whether GPT-5.6 launch-week load contributed to the incident — that remains inference, not a stated cause. We also do not know when the separate FedRAMP workspace issues will be fully closed.
FAQ
Is ChatGPT down right now?
As of the July 12, 2026 incident, ChatGPT was reported operational again by 9:46 AM UTC after a brief iOS 403 outage that began around 5:29 AM ET. The fastest way to confirm current status is OpenAI’s official status page at status.openai.com, which shows live incident updates.
What does a 403 error mean in ChatGPT?
A 403 error means the server understood the request but refused to authorize it — “forbidden.” On July 12, iOS users saw 403 errors when loading or continuing conversations, so chats would not open or respond even though the app itself launched. These errors usually clear once the underlying service issue is resolved.
Was the July 12 ChatGPT outage related to the FedRAMP issues?
No. OpenAI’s status page treated them as separate incidents. The iOS 403 errors were a brief consumer outage that resolved within hours, while the FedRAMP workspace issues — affecting Codex, workspace analytics, conversation search and more — had been ongoing for about a week and target enterprise government-grade workspaces.
What should I do when ChatGPT shows a 403 error?
For a server-side 403 like the July 12 incident, there is little a user can do except wait, since the issue is on OpenAI’s end rather than your device. Check status.openai.com to confirm it is a known outage, and retry once the status page shows the incident resolved.
