OpenAI Outage History

OpenAI reported a brief degraded performance in ChatGPT responses, with elevated latency impacting the ChatGPT / Conversations component. The incident began at Feb 20, 2026 10:01 PM UTC when OpenAI flagged elevated latency. A follow-up update at Feb 20, 2026 10:26 PM UTC noted that latency was increased for some users and that investigators were looking into it. The issue was resolved by Feb 20, 2026 10:32 PM UTC, returning the ChatGPT / Conversations service to normal. The provider issued 3 status updates during the incident, and the outage lasted approximately 31 minutes. This appears as a brief degradation rather than a full outage.

The OpenAI Sora 2 video generation service experienced degraded performance, impacting users attempting to generate videos. The issue began at Feb 18, 2026 3:36 PM UTC and was resolved by Feb 18, 2026 4:31 PM UTC after engineering mitigations and investigation, lasting about 55 minutes. Users reported failures such as videos not being produced and prompts returning errors.

OpenAI issued three status updates during the incident. The first stated degraded performance for Sora 2, the second indicated ongoing investigation, and the final update confirmed that the service had returned to normal. The activity appeared most heavily in North America, with reports from the United States and Canada indicating the scope of the impact.

OpenAI's Shopping Research feature within ChatGPT / Search went offline starting at 8:51 PM UTC, blocking access to the shopping research capability for users. A small number of user reports were observed during the incident.

Mitigation steps were implemented at 8:57 PM UTC and the Shopping Research feature returned to normal by 9:17 PM UTC, with the service bringing back full functionality. OpenAI issued three status updates during the incident as teams investigated and restored the feature.

OpenAI's Codex Cloud experienced elevated error rates, causing degraded performance for ChatGPT / Codex features for a brief period. User reports appeared around 5:15 PM UTC, and the Codex component was listed as degraded at Feb 16, 2026 5:31 PM UTC. OpenAI issued 3 status updates, applied mitigation, and restored normal service by Feb 16, 2026 5:51 PM UTC, for an outage duration of about 20 minutes. The incident affected the Codex-related components and was resolved with users returning to normal operation after the mitigation.

OpenAI reported degraded image generation starting at Feb 14, 2026 1:36 AM UTC, affecting two components: ChatGPT / Image Generation and APIs / Images. Users encountered high error rates, with some experiencing server not responding or heavy load messages while attempting image generation. The degraded performance persisted for about 2 hours and 50 minutes and was resolved by Feb 14, 2026 4:26 AM UTC.

OpenAI's image generation services experienced a brief degradation affecting ChatGPT / Image Generation and the APIs / Images components. Users may have encountered elevated errors or slower image generation during the incident. The outage began at Feb 13, 2026 8:17 AM UTC and was resolved by Feb 13, 2026 9:01 AM UTC, lasting about 44 minutes. User reports around 8:30 AM UTC indicated issues with image generation.

OpenAI posted updates on the status page: the initial notice at Feb 13, 2026 8:17 AM UTC warned of high errors with image generation and noted that mitigation was underway, followed by a resolution update at Feb 13, 2026 9:01 AM UTC indicating services had returned to normal. The incident affected two components: ChatGPT / Image Generation and APIs / Images, and involved two official status updates during the event.

OpenAI reported intermittent issues affecting ChatGPT Conversations, causing users to experience chat interruptions. The incident began on Feb 12, 2026 9:41 PM UTC and was resolved by Feb 12, 2026 9:57 PM UTC, lasting roughly 16 minutes. User reports surged around 9:15 PM UTC, peaking at 21:15 UTC as some users reported messages not sending or delayed responses. The issue was isolated to the ChatGPT Conversations component during this window.

OpenAI's status page documented the incident, with two official updates describing the investigation and subsequent resolution. The problem affected a subset of users engaging in conversations with ChatGPT, but service was restored shortly after.

OpenAI reported a degraded state affecting ChatGPT Conversations and the text-embedding-3-small API beginning at Feb 11, 2026 4:31 PM UTC. Users experienced issues starting conversations in ChatGPT and elevated errors plus latency for embedding requests across regions, with a noticeable impact on response times and overall usability. The incident spanned across two OpenAI components and was observed by a broad set of users worldwide, prompting ongoing investigation and monitoring.

OpenAI experienced a brief degradation in ChatGPT login, with users reporting sign-in errors and difficulties accessing the service. The outage began at Feb 10, 2026 10:41 PM UTC and was resolved by Feb 11, 2026 12:42 AM UTC, lasting about 2 hours and 1 minute. Access issues affected ChatGPT login across platforms for some users.

The OpenAI status page documented the incident, initially noting increased login errors for ChatGPT and later confirming a return to normal. The status page issued 2 updates during the incident. User reports were limited, but included examples of login difficulties from a subset of users.

OpenAI reported a brief degraded performance for ChatGPT Go affecting the ChatGPT / Conversations component. Users attempting to use ChatGPT Go may have encountered elevated errors and intermittent failures. The outage began at Feb 10, 2026 8:11 PM UTC and was resolved by Feb 10, 2026 9:02 PM UTC, lasting about 51 minutes. OpenAI issued 2 status updates during the incident to communicate the issue and its resolution. Reports were limited and did not show a clear regional pattern, indicating a localized impact rather than a global outage.