ChatGPT, Claude Outages Raise Reliability Concerns

ChatGPT drew nearly 1,000 outage reports on September 29, the same day Anthropic’s Claude logged more than 11,000, raising reliability questions about major generative AI services. Nearly 1,000 outage reports related to OpenAI were filed with the tracking site Downdetector around 2 p.m. Eastern time on September 29, with about 85% of the reported problems involving ChatGPT use. The simultaneous disruptions underscore persistent operational challenges as these AI platforms expand their enterprise reach.

ChatGPT Experiences Intermittent Service Problems

OpenAI said on its official status page that ChatGPT was experiencing higher-than-usual error rates, and that some users might encounter failed requests, problems logging in or signing up, and tasks that do not run to completion. Errors were also reported in the coding agent Codex and parts of the API.

The scale of the disruption swung sharply by the hour. Downdetector reports fell below 100 around 2:30 p.m. before climbing back above 500 roughly 15 minutes later, according to Forbes. That pattern suggests intermittent problems recurring in certain regions and services rather than a full outage locking out all users at once.

The disruption came as OpenAI held its annual developer conference the same day. At the event, the company unveiled Dots, a new AI agent that carries out tasks continuously on a user’s behalf.

Claude Suffers Larger Wave of Outage Reports

Anthropic’s AI chatbot Claude also went down the same day. The service issues affected multiple Claude products, including the main platform, Claude Code, and parts of the company’s API infrastructure. According to the provided sources, Claude’s outage drew significantly more user reports than ChatGPT’s disruption.

The timing of the dual outages has prompted questions about infrastructure dependencies and the reliability of AI services that are increasingly embedded in enterprise workflows. Both companies have experienced multiple service disruptions throughout 2026, though the exact causes of the September 29 incidents were not immediately disclosed by either organization.

Recovery and Ongoing Reliability Concerns

Both services eventually recovered from their respective outages. The fluctuating pattern of reported issues suggests that the problems may have been related to specific geographic regions or service components rather than complete system failures.

The incidents come as both OpenAI and Anthropic are expanding their enterprise offerings. Barclays plans for half of its software developers to use Claude Code by the end of 2026 as it modernizes legacy systems and improves operational efficiency. Anthropic made Claude for Government generally available to U.S. federal and state agencies, following an earlier beta period.

As businesses increasingly rely on AI chatbots for critical operations, the simultaneous service disruptions highlight potential vulnerabilities in enterprise AI infrastructure. The recurring nature of these outages throughout 2026 raises questions about scalability and redundancy planning as demand for generative AI services continues to accelerate.

Key Facts

ChatGPT received nearly 1,000 outage reports on September 29 around 2 p.m. Eastern time
Claude logged more than 11,000 outage reports the same day
Approximately 85% of OpenAI-related reports involved ChatGPT service issues
OpenAI held its annual developer conference on September 29, unveiling the Dots AI agent
Service issues affected ChatGPT, Codex, Claude, Claude Code, and API infrastructure for both platforms
Downdetector reports for ChatGPT fluctuated between below 100 and above 500 within a 15-minute period