OpenAI faces second major outage in July, ChatGPT services disrupted globally
The outage, which began around 6:10 AM IST on July 16, marks the second significant service failure this month, sparking growing concerns about the platform’s reliability as demand for artificial intelligence tools surges globally.
ADVERTISEMENT
OpenAI experienced a global outage that brought its popular AI services including ChatGPT, Sora and the GPT API to a halt for millions of users worldwide. The outage, which began around 6:10 AM IST on July 16, marks the second significant service failure this month, sparking growing concerns about the platform’s reliability as demand for artificial intelligence tools surges globally.
According to outage tracking platform Downdetector, over 88% of user complaints were related to ChatGPT being entirely inaccessible. Others reported errors with OpenAI’s video-generation tool Sora and developer platform Codex.
Users from North America, Europe and Asia, including India, flooded social media platforms with screenshots and messages detailing issues such as login failures, incomplete responses, sudden disconnections and verification loops.
OpenAI acknowledged the issue through its official status page, confirming “degraded performance” across multiple services. “We have identified elevated error rates and are working on a mitigation,” the company said, without specifying a timeline for resolution or the root cause of the outage. It remains unclear whether the disruption stems from a server overload, a software glitch, or another technical failure.
The latest breakdown has disrupted workflows for developers, content creators, researchers and businesses that rely on OpenAI’s tools daily.
Many voiced their frustration online, questioning the platform’s stability amid increasing enterprise adoption. OpenAI has advised users to refrain from repeated login attempts, which could trigger security locks and to monitor the status page for real-time updates. It also recommended saving work externally to avoid potential data loss.
