Millions of users worldwide faced a significant disruption to their workflows early Wednesday morning as ChatGPT suffered a major service outage.
The incident, which began shortly before 6:30 AM, rendered the popular AI chatbot inaccessible for many and caused alarming data visibility issues where conversation histories appeared to vanish completely. OpenAI has since confirmed that all systems are fully operational following a rapid mitigation effort.
According to OpenAI’s official status page, the incident was first identified as affecting the creation of new Codex tasks, a critical backend component powering code generation and reasoning features.
Engineering teams moved quickly to address the problem, updating the status to “Monitoring” at 06:29 AM after applying a fix. By 07:16 AM, the company declared the incident resolved, stating that all impacted services had fully recovered.
This 45-minute window of critical instability highlights the fragility of cloud-based AI dependencies for global enterprise users.
Missing History Conversations
While the official root cause was technical, the user-facing symptoms were particularly distressing. Unlike typical connection timeouts, this outage manifested as a “missing history” bug, where the sidebar containing past chats and saved conversations failed to load entirely.
This specific error is a recurring source of anxiety for power users who rely on ChatGPT as a repository for code snippets, creative drafts, and research logs.
Social media platforms saw a spike in reports from users fearing their data had been permanently deleted, though this behavior is often a temporary side effect of backend API failures rather than actual data loss.
The specific identification of “Codex tasks” as the culprit suggests this outage disproportionately affected developer workflows and API integrations involving code generation.
Codex is the engine responsible for translating natural language into code, meaning that during the downtime, automated coding assistants and third-party applications relying on these specific API endpoints likely failed silently or returned errors.
Although the service has been restored, OpenAI advises users to refresh their sessions to ensure full functionality returns.
Follow us on Google News, LinkedIn, and X for daily cybersecurity updates. Contact us to feature your stories.
