OpenAI said some users were facing elevated error rates on ChatGPT and Codex on Tuesday, as an ongoing Chatgpt Outage stretched into its fourth hour. The company said it had applied a mitigation about two hours earlier and was now monitoring the recovery.
The incident is the kind that sends people straight to the status page because it does not read like a full shutdown, only a partial failure that can still lock out anyone trying to work through the two services. OpenAI’s wording was plain: "We are investigating the issue for the listed services."
That is why the search traffic moved now. Reports on DownDetector showed sharp spikes over the past few hours, and users on X were still saying the problem had not gone away even after OpenAI said it had applied a fix. The company has not said what caused the error rates, only that it was continuing to watch recovery.
The gap between mitigation and relief matters. OpenAI can say it has taken action, but the live reports suggest the outage was not yet behaving like a resolved incident, which leaves the company in a familiar place: trying to calm users while the services themselves are still unstable.
For Marcus Mendes, who covered Apple news in Brazilian media in 2012 before widening his focus to the broader tech industry, the update is the part readers care about most: ChatGPT and Codex were still listed as affected, and OpenAI was still stopping short of declaring the issue over.
