ChatGPT experienced a global outage on July 25, 2026, starting around 4:47 AM ET, with users worldwide unable to load chats or send prompts across web, mobile, API and Codex services. OpenAI acknowledged the issue on its status page and applied a fix, restoring service within about 50 minutes. This marked the second major outage to hit ChatGPT in just two days, following a nearly 24-hour disruption on July 23-24, 2026, raising fresh questions about reliability for the millions of users and businesses now dependent on the platform.
- ChatGPT suffered a global outage starting around 4:47 AM ET on July 25, 2026, affecting web, mobile, API and Codex services.
- Users reported stalled prompts, failure to load chat history, and 'too many concurrent requests' errors.
- OpenAI acknowledged the issue and applied a fix, with services fully recovered within about 50 minutes.
- This was the second major ChatGPT outage in two days, following a nearly 24-hour disruption on July 23-24.
- As of the latest checks, ChatGPT is reported operational with no ongoing outage.
ChatGPT Goes Down Worldwide
OpenAI's ChatGPT experienced a significant global outage on Saturday, July 25, 2026, leaving thousands of users across the US, India, Europe and other regions unable to access the service. Independent outage tracker Downdetector registered a sharp surge in complaints starting at around 4:47 AM Eastern Time, with users reporting stalled prompts, failure to load previous chat history, and persistent connectivity errors on both the web interface and mobile apps.
What Users Experienced
Affected users described ChatGPT getting stuck on loading animations in the sidebar, alongside "too many concurrent requests" error messages when attempting to send messages. The disruption was not limited to casual users chatting on the free or Plus tiers; enterprise subscribers and developers relying on API integrations were also hit. OpenAI's coding platform, Codex, was affected as well, with the company's status page listing as many as 12 API endpoints experiencing issues.
OpenAI's Response and Recovery
OpenAI acknowledged the problem on its official status page at around 5:30 AM ET, stating it was "investigating the issue for the listed services." The company applied a fix and monitored the situation as reports continued to come in. According to updates, ChatGPT had fully recovered by around 6 AM ET, meaning the outage lasted approximately 50 minutes from detection to resolution. Frustrated users had taken to social media in the interim, noting that the outage disrupted their work, coding projects and writing tasks.
Second Outage in Two Days
Notably, this was not an isolated incident. A separate, more prolonged outage had already hit ChatGPT and its associated services just a day earlier, beginning on July 23 and stretching for nearly 24 hours before OpenAI resolved it by the afternoon of July 24. That earlier disruption similarly affected ChatGPT, Codex and OpenAI's APIs, with the company's status page showing an elevated rate of errors throughout. Taken together, the two incidents mark a rough stretch of reliability issues for OpenAI's flagship product within a single week.
Why This Matters
With ChatGPT now deeply embedded in workflows ranging from coding and content creation to customer support and research, even short outages can have an outsized impact on productivity for millions of users and businesses that depend on it daily. The back-to-back disruptions have reignited discussion among users and developers about the risks of over-reliance on a single AI service provider for critical tasks, and the importance of building in fallback options where possible.
Current Status
As of the latest checks, ChatGPT's status pages and independent monitors indicate the service is operational, with no ongoing official outage reported. Users experiencing continued issues are advised to check OpenAI's official status page for real-time updates.
Comments (0)
Leave a Comment
No comments yet. Be the first to comment!