A fresh wave of technical trouble has hit ChatGPT users, with OpenAI confirming degraded performance and widespread outages on its platform, particularly for mobile users.
The disruption, which began around 2:05 PM UTC on April 2, 2025, has affected users across Android and iOS, although the web version remains mostly stable.
According to OpenAI’s official status page, the company is actively investigating increased error rates and slower performance in real time.
The outage comes at a time when usage of ChatGPT is surging globally, especially with added functionalities like image generation and voice synthesis, prompting speculation around server overload as a possible trigger.
Error reports jump 23,000%
OpenAI’s own status portal indicates that the outage centres on ChatGPT’s core functionality, with three internal components currently disrupted.
In the last 24 hours, Downdetector logged 1,642 issue reports from users—an enormous increase from the platform’s typical daily average of just seven complaints.
The issue was officially acknowledged at 7:40 PM IST, five minutes after users began raising alarms online and through OpenAI’s feedback channels.
The company said it is still investigating the root cause, with no timeline yet offered for full service restoration.
ChatGPT’s support portal has been directing users to report issues not only with the chatbot, but also with other OpenAI services such as Dall-E and Sora.
However, OpenAI has stated that the current disruptions are mainly affecting ChatGPT’s user-facing performance.
Web stable, apps disrupted
Despite the growing number of complaints, ChatGPT’s browser-based version is reportedly functioning normally.
The issue appears to be localised to mobile platforms, with users encountering error messages such as “Error in message stream” when attempting to use the chatbot.
Platforms like X (formerly Twitter) and Reddit have seen a flood of user posts describing failed attempts to generate responses or access ChatGPT’s new tools.
Many reports are tied to mobile-specific access methods, indicating that the infrastructure for app-based interactions may be under strain.
OpenAI has yet to clarify whether the issue lies in its servers, network capacity, or recent feature rollouts.
However, the company’s communication suggests ongoing monitoring and a commitment to regular updates as engineers continue their investigation.
New features may be the cause
Although no official cause has been confirmed by OpenAI, user speculation and timing suggest that recently added features—particularly Ghibli-style image generation and real-time voice capabilities—may be putting new pressure on backend systems.
The AI tool has seen significant upgrades in recent weeks, including the expansion of ChatGPT’s multimodal functionalities.
These updates have likely led to a higher volume of requests and more complex data processing, which may be contributing to server stress and resulting errors.
Notably, similar issues were not reported in other OpenAI tools like Dall-E or Sora at the time of the incident, further pointing to specific challenges within the ChatGPT framework.
The post Is ChatGPT down? OpenAI confirms outage, investigating errors appeared first on Invezz