ChatGPT Down: Thousands Report Issues With Login And Chat Across Web And App

On July 25, 2026, ChatGPT, its app, and APIs suffered a major global outage with thousands of reports on Downdetector. OpenAI logged "elevated error rates" and applied a fix, starting recovery.

ChatGPT, its app, and APIs suffered a major global outage with thousands of reports on Downdetector.

Source : X

Show Quick Read

Key points generated by AI, verified by newsroom

• OpenAI services, including ChatGPT, suffered widespread outage.
• Disruption began July 25, 2026, impacting web and apps.
• Downdetector showed sharp spike; OpenAI confirmed investigating issues.
• Company applied mitigation, currently monitoring recovery efforts.

Thousands of users were left unable to access OpenAI services on the afternoon of July 25, 2026, as ChatGPT experienced a sudden, widespread outage affecting both its web interface and mobile applications.

The disruption began around 2:54 PM, with users reporting failed logins, stuck chats, and complete loss of access across multiple regions.

Sudden Spike on Downdetector

Independent monitoring service Downdetector recorded a dramatic surge in user reports within minutes of the initial failure:

• Baseline Reports: 2 (normal operation)

• Peak Surge: 1,535 reports within minutes

• Affected Platforms: ChatGPT Web, iOS/Android Apps, and developer APIs

The steep curve indicated a global server-side issue rather than localized network friction or individual account glitches.

Now the spike is slowly making a negative curve as till 3.24 pm, the rotal reports were 365. 

In the US, there were 3,243 reports at the time of writing, with 80% citing the web version, 8% the app and 8% Codex. Meanwhile, in the UK, there were 3,088 reports, with 80% citing ChatGPT, 9% citing the app, and 8% citing Codex.

OpenAI Confirms Elevated Error Rates

According to OpenAI’s status page, the issue has been logged under “Elevated error rates," with the company stating it is actively investigating the problem across the listed services.

There have acknowledged the disruption on its status page, stating, "We're currently experiencing issues. We have applied the mitigation and are monitoring the recovery."