OpenAI Investigates Elevated Errors for ChatGPT Plus and Pro Users

OpenAI posted an incident notice on its official status page on September 22, 2026, reporting elevated error rates for ChatGPT Plus and Pro users, naming ChatGPT’s Conversations component as the single affected component and stating that it was working to implement a mitigation.

The status page records the incident under the title “Increased error rate for Plus and Pro users.” It was first posted at 9:58 a.m. UTC, shown as 5:58 a.m. on the page, and is classified as Degraded performance while in the Investigating state. The record is also listed on OpenAI’s status history page under September 22 with the same 5:58 a.m. timestamp.

In the notice’s only update, OpenAI said it had “identified that users are experiencing elevated errors for the impacted services” and was working on implementing a mitigation. The incident’s affected-components section lists the ChatGPT product group with one affected component, Conversations, while the title scopes the impact to users on the Plus and Pro subscription tiers.

OpenAI’s main status page displayed a banner warning that it was currently experiencing issues and listed the incident under ChatGPT as ongoing and affecting ChatGPT.

Availability Metrics on OpenAI’s Status Page

The system status section of the status page, covering June 2026 through September 2026, reports the ChatGPT group at 99.65% uptime across 15 tracked components. Conversations, the component named in the new incident, is listed at 99.96%. The page lists Login, Codex in ChatGPT Desktop, Compliance API, Search, File uploads, GPTs, Deep Research, Agent, ChatGPT Atlas and Sites at 99.98% each, with ChatGPT Work at 99.84%, Voice mode at 99.96%, Connectors/Apps at 99.96% and Image Generation at 99.79%.

The APIs group shows 12 components at 99.93% uptime, with Images at 99.95%, Login at 99.995% and Sora at 99.97%, while Chat Completions, Responses, Fine-tuning, Embeddings, Batch, Audio, Moderations, Realtime and Files are each listed at 100%. The Codex group lists its four components — Codex Web, Codex API, CLI and the VS Code extension — at 100% uptime, and the single FedRAMP component is listed at 100%. The Ads Platform section lists Ads Manager and the Ads API, with the Ads API shown at 100%.

OpenAI notes on the page that availability metrics are reported at an aggregate level across all tiers, models and error types, and that individual customer availability may vary depending on subscription tier and the specific model and API features in use.

Earlier September Incidents

The history page lists a series of resolved incidents from earlier in September 2026, including one affecting the same component and the same paid tiers. On September 9, 2026, an incident titled Increased Error Rate For Pro and Plus Plan Conversations listed Conversations as its only affected component. It opened with an Investigating update at 10:51 a.m. using the same wording as the September 22 notice, moved to Monitoring at 11:44 a.m. after OpenAI said it had applied a mitigation, and was marked Resolved at 12:21 p.m., with OpenAI stating that all impacted services had fully recovered.

On September 17, 2026, an incident reporting elevated error rates across API models listed 12 affected API components, including Chat Completions, Responses, Realtime and Sora; it was identified at 4:20 p.m., moved to Monitoring at 5:20 p.m. after a mitigation was applied, and was resolved at 5:50 p.m. On September 3, 2026, an incident reporting elevated errors across ChatGPT and Codex listed 15 affected ChatGPT components and four affected Codex components; it was resolved at 12:55 p.m., and OpenAI later published a write-up, with a resolution note stating that some Codex remote control users might need to pair their mobile device again.

Other resolved entries on the September history include elevated errors for ChatGPT users in Europe on September 11 and 13, 2026, elevated errors from GPT-5.6 and GPT-5.6 Instant on paid plans on September 15, 2026, elevated errors in ChatGPT Work on September 2, 10, 14 and 16, 2026, SSO sign-in and SCIM provisioning issues on September 18, 2026, overbilling for OpenAI-hosted containers in the Agent API on September 19, 2026, and increased errors for users in the APAC region across ChatGPT, Work, image generation, file upload, Voice and Codex Cloud on September 4, 2026.

As of 12 minutes after the initial notice, the Investigating update remained the only entry on the September 22 incident record, and OpenAI had not posted a notice that a mitigation had been applied or that impacted services had recovered.