On Monday morning, OpenAI’s flagship chatbot, ChatGPT, experienced a significant service interruption, leaving users unable to access their conversation histories or receive timely responses. The instability, which peaked around mid-morning, underscored the growing, yet fragile, reliance on large language models for daily productivity.
Reports on the monitoring site Downdetector surged as users noticed the left-hand sidebar—typically populated with past interactions—had vanished. Beyond the interface issues, the underlying engine struggled to process prompts, with many requests timing out or hanging indefinitely. OpenAI later confirmed the outage on its official status page, noting that the disruption also extended to Codex, the company’s specialized model for programming tasks.
As artificial intelligence shifts from a novelty to a fundamental utility, these moments of downtime serve as a reminder of the centralized infrastructure supporting the current AI landscape. For the thousands of developers and professionals who now integrate these tools into their daily workflows, a brief outage is no longer a minor inconvenience, but a genuine friction in the machinery of modern work.
With reporting from [Canaltech].
Source · Canaltech
