OpenAI’s ChatGPT experienced a significant partial outage on April 20, 2026, leaving thousands of users globally unable to access the world’s most widely used AI chatbot for approximately 90 minutes. The issues first started at around 10:05 AM Eastern Time, when a large spike in reports appeared on Downdetector. At their peak, there were over 8,700 reports of issues in the United Kingdom and just over 1,900 in the United States. The OpenAI status page confirmed there was a partial outage for at least 90 minutes before a fix was deployed.
OpenAI’s status page showed a partial outage, with the company stating it was investigating issues affecting ChatGPT, Codex, and the API Platform, with impacted users unable to access these services. Just about all parts of ChatGPT were experiencing issues simultaneously, ranging from conversations and login to voice mode and image generation. Users reported a range of symptoms, from blank pages on the website to long response times, with those already logged in largely seeing problems accessing existing conversations while others were locked out entirely at the login stage. The two services, ChatGPT and Codex, experiencing downtime together indicated that their systems encountered a backend or server problem rather than an issue isolated to a single product.
OpenAI posted an update at 12:48 PM Eastern Time saying it was monitoring the recovery following a fix, and Downdetector reports returned to baseline levels shortly after, suggesting the outage was effectively resolved for the vast majority of users within a roughly 90-minute window. The main issue for most affected users was accessing old conversations, accounting for approximately 63 percent of reported problems, compared to 27 percent who were unable to sign in at all.
The outage drew significant attention on social media, with users expressing frustration at how central ChatGPT has become to daily work and productivity. OpenAI’s services are used by millions of people directly through ChatGPT and by tens of thousands of businesses and developers through the API, meaning an outage of this scale affects not just individual users but downstream products and services built on the OpenAI platform. The incident serves as a reminder of the infrastructure dependency that has grown around AI services in a relatively short period, with even a brief disruption capable of disrupting workflows across industries and geographies simultaneously.
Follow the SPIN IDG WhatsApp Channel for updates across the Smart Pakistan Insights Network covering all of Pakistan’s technology ecosystem.
