AI Models & Platforms

OpenAI Confirms Service Degradation Hitting ChatGPT and Codex Users

mm
Published
Add Unite.AI to your preferred sources on Google

OpenAI opened an incident on its official status page on September 3, 2026, after detecting elevated errors across ChatGPT and Codex, with the incident record listing degraded performance in every tracked component of both services.

The record carries the title “Elevated errors across ChatGPT and Codex” and classifies the event as “Degraded performance” with an “Investigating” status. OpenAI published its first and, as of the latest check, only update at 2:58 PM, stating: “We are investigating the issue for the listed services.” The affected-components window on the page opened at 2:56 PM.

Nineteen Components Degraded Across Two Service Groups

The incident lists 19 degraded components spread across the ChatGPT and Codex groups on OpenAI’s status page. The ChatGPT group accounts for 15 affected components: Conversations, Login, ChatGPT Work, Codex in ChatGPT Desktop, Compliance API, Search, File uploads, Voice mode, GPTs, Image Generation, Deep Research, Agent, ChatGPT Atlas, Sites, and Connectors/Apps.

The Codex group accounts for the remaining four: Codex Web, Codex API, the Codex command-line interface, and the Codex VS Code extension. The incident record marks every one of these components as experiencing degraded performance.

APIs, FedRAMP, and Ads Platform Not Listed

The incident does not extend to OpenAI’s other tracked service groups. The record lists no affected components in the APIs group, which covers 12 components including Chat Completions, Responses, Fine-tuning, Embeddings, Images, Batch, Audio, Moderations, Realtime, Files, Login, and Sora. The single-component FedRAMP group and the two-component Ads Platform group, comprising Ads Manager and Ads API, also appear nowhere in the incident’s affected-component list.

The main OpenAI status page carried a banner reading “We’re currently experiencing issues” while the incident remained open, naming ChatGPT and Codex as the affected services. Its rolling availability figures showed 99.94 percent uptime for APIs, 99.62 percent for ChatGPT, 99.98 percent for Codex, and 100 percent for FedRAMP over the June 2026 to September 2026 window. The page notes 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.

Second ChatGPT Incident of the Day

The elevated-errors incident is the second entry OpenAI logged for ChatGPT on September 3, 2026. Its status history index also records a “ChatGPT Work Mode High Error Rates” incident that was resolved at 12:10 AM that morning with the note that all impacted services had fully recovered. The record for that incident shows OpenAI opened an investigation at 12:04 AM, six minutes before declaring recovery, and classified it as a partial outage affecting the ChatGPT group.

The history index lists a third recent event the prior day, September 2, 2026, when an “Elevated errors creating new accounts” incident was marked resolved with the note that the issue affecting new account creation had been resolved.

How OpenAI Structures Service Reporting

OpenAI organizes its public status reporting into five groups: APIs, ChatGPT, Codex, FedRAMP, and Ads Platform. The ChatGPT group tracks the consumer and workspace product surface, from core conversation handling and sign-in to voice, search, file uploads, and newer additions such as the ChatGPT Atlas browser, the Agent feature, Sites, and ChatGPT Work. The Codex group covers the coding assistant’s surfaces: the web interface, the API, the terminal-based CLI, and the VS Code editor extension.

The current incident touches the full breadth of both groups. On the ChatGPT side it spans entry points such as Login, core functions such as Conversations and Voice mode, and workspace features such as ChatGPT Work and Connectors/Apps. On the Codex side it covers every environment in which the coding agent runs, including the browser-based Codex Web, the developer-facing Codex API, the local CLI, and the editor extension, as well as the Codex in ChatGPT Desktop component tracked inside the ChatGPT group.

OpenAI had not named a root cause in the incident record as of the latest published update, and the page lists no mitigation or resolution entries beyond the initial investigation notice. The status page offers a subscription option for updates, and the incident record remains the company’s designated channel for further notices on the event.

mm

Jonas Reeve is an AI-generated analyst at Unite.AI, focusing on cognitive AI, artificial general intelligence (AGI), and the theoretical foundations of machine intelligence. His work explores how learning, reasoning, memory, and abstraction emerge in both biological and artificial systems, drawing connections between modern AI architectures and long-standing questions in cognitive science and philosophy of mind.

With a conceptual and reflective approach, Jonas examines frameworks such as reasoning models, agentic systems, emergent cognition, and alignment theory, aiming to clarify what progress toward AGI actually means—and what it does not. Rather than chasing timelines or hype, he emphasizes first principles, conceptual rigor, and the limits of current models.

Articles authored by Jonas Reeve are AI-generated and reviewed by Unite.AI’s editorial team to ensure accuracy, clarity, and responsible discussion of advanced AI concepts.