Liputan6.com, Jakarta - The popular artificial intelligence service, ChatGPT, experienced a major outage on Tuesday, February 3, 2026, leaving thousands of users worldwide unable to access the platform.
The outage began around 3:00 PM ET (Eastern Time), with reports of issues surging sharply within just 20 minutes.
Advertisement
OpenAI, the company that developed ChatGPT, quickly confirmed the issue with its service.
The impact of the outage was particularly felt in the United States, where outage tracking site Downdetector recorded a surge in reports from dozens to over 13,000.
Users worldwide reported error messages such as "Hmm... something seems to have gone wrong" or "An error occurred" when trying to use the platform.
ChatGPT Global Outage: How It Happened and Impact
On February 3, 2026, ChatGPT experienced a service outage affecting over 12,000 users globally.
Initial reports surfaced around 3:00 PM ET, and within a short time, complaints skyrocketed, peaking at over 10,000 to 12,000 reported cases on Downdetector.
The outage wasn't limited to a single feature; it affected nearly every aspect of the ChatGPT platform.
Features such as conversation history, image generation via DALL-E, search functionality, and the newly launched Codex tool were all affected.
OpenAI acknowledged a "high error rate" for both ChatGPT users and the platform in general, including a "high error rate for fine-tuning work."
OpenAI's Rapid Response and Service Restoration
Soon after the outage began, OpenAI responded by acknowledging the issue via its official status page.
Some analysts speculated that a surge in new users and API calls from new integrations, such as the native Codex app for macOS and Apple's Xcode update, may have overloaded the existing infrastructure.
At 5:14 PM ET, OpenAI marked the issue related to "high error rates" for ChatGPT and Platform users as resolved.
OpenAI said on its status page that it had applied the "necessary mitigations," and the page shows ChatGPT as "operational" as of 6:30 p.m. ET.
The service restoration process was gradual, with users in the United States seeing a rapid decrease in errors around 4:00 PM ET, followed by restoration in other international regions.
OpenAI Strategy Change: Deprecation of Legacy Models and WhatsApp Service
In addition to the disruption incident, OpenAI also announced several strategic changes related to its products.
Starting February 13, 2026, OpenAI will deprecate several legacy ChatGPT models, including GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini.
This decision aims to simplify the model catalog and focus resources on developing newer, more efficient models, such as the GPT-5 series.
This move is based on usage data showing that only about 0.1% of ChatGPT users were still using GPT-4 daily at the time of the announcement.
This indicates that most users have shifted to the more advanced and widely used GPT-5 series models. This deprecation is expected to improve the performance and innovation of OpenAI's flagship models.
Another change is the discontinuation of the ChatGPT chatbot service on WhatsApp, which will officially take effect on January 15, 2026.
This decision was made due to a change in WhatsApp policy prohibiting third-party AI services on Meta's messaging platform.