A major global outage on June 10, 2025, left millions of ChatGPT users unable to access the service for over 12 hours. The outage affected both free and paid tiers as well as OpenAI’s API and image generator, Sora.
A global outage brought OpenAI’s ChatGPT to a halt on June 10th, 2025. The incident affected millions of users across continents. The disruption, which started around 11 a.m. GMT, impacted both the ChatGPT app and its developer API. The outage also extended to OpenAI’s image generation tool, Sora. Users reported error messages, failed loads, and significant latency.
Reports on Down Detector spiked early in the morning and surged again as more users in Europe, North America, and Australia came online. By 8:23 p.m. GMT, there were over 500 reported issues, suggesting that the outage had a lasting impact.
Read also:
Within the first two hours of the disruption, OpenAI confirmed that it had identified the root cause. Despite this, services remained unreliable for approximately 10 hours. Some users saw the chatbot return vague error messages like, “Hmmm… something seems to have gone wrong,” while others experienced total failure to load the platform.
This isn’t the first time a high-traffic AI system has experienced problems, but the scale of ChatGPT’s user base, over 500 million globally, makes this incident particularly significant.
Experts in the tech space were quick to weigh in. Bob Wambach, VP of Portfolio and Strategy at Dynatrace, commented, “The complexity of the IT systems underpinning public-facing AI services, combined with rapidly increasing interest and user traffic, makes ensuring reliability and resilience a challenge.”
He added that relying on human capacity alone to maintain real-time oversight of such systems is no longer feasible, stating, “Only AI can monitor, analyze, and optimize these digital environments at the speed and scale required.”
Here’s a timeline of how the global ChatGPT outage unfolded, as reported by Toms Guide. Each timestamp captures a major development in OpenAI’s response and user experience throughout the day.
06:36 GMT: Outage begins
OpenAI first acknowledged a disruption on its system status page, citing “elevated error rates and latency” affecting ChatGPT’s web, mobile, API, and Sora services.
07:45 GMT: Early reports surge
Outage reports began appearing on Down Detector, peaking above 1,000 user submissions as the issue gained attention.
10:32 GMT: Partial outage confirmed
OpenAI officially labeled the incident a "partial outage" and confirmed that it was actively investigating the problem.
10:47 GMT: ‘Too Many Requests’ errors
Users began encountering “Too many concurrent requests” notifications and significant slowdowns in response times.
11:02 GMT: System component impact listed
The OpenAI status page showed that 21 ChatGPT components, 14 API components, and 4 Sora components were affected.
11:11 GMT: Users flock to alternatives
Google Trends data showed spikes of over 90% in searches for ChatGPT alternatives like Claude and DeepSeek.
11:40 GMT: Second spike
A second surge in outage reports hit Down Detector as users in Europe, North America, and Australia came online.
11:55 GMT: Global reach confirmed
OpenAI’s paid and free tiers were confirmed to be affected globally, with users experiencing errors like “A network error has occurred” and “Error in message stream.”
12:24 GMT: Root cause identified
OpenAI stated it had pinpointed the underlying issue and was implementing a fix.
12:29 GMT: Status turns yellow
The system status indicator turned from red to yellow, signaling partial restoration as mitigations took effect.
17:49 GMT: Services restored for some
Users in regions including the US and the Philippines reported access had returned, though performance remained inconsistent. Down Detector reports fell below 600.
18:06 GMT: OpenAI monitoring fix
OpenAI posted: “We are still monitoring the fix and working towards full recovery,” while user complaints persisted but declined.
18:44 GMT: Outage nears resolution
OpenAI marked most systems as operational, with ChatGPT and Sora stabilizing and fewer new reports coming in.
As AI becomes more deeply integrated into both consumer and enterprise settings, outages like this reveal critical vulnerabilities. For individuals, it means interrupted conversations, missed deadlines, or lost productivity. For businesses, it could mean system failures, customer service delays, and significant financial consequences.
With OpenAI valued at $300 billion and regarded as a global leader in AI development, any major service disruption raises questions about reliability, infrastructure readiness, and the growing demand for stability in AI services.
This ChatGPT outage is a reminder of the world’s increasing dependence on artificial intelligence (AI). While OpenAI has made progress toward restoring services, the situation indicates the need for scalable, resilient infrastructure capable of supporting the demands of global AI usage.
OpenAI continues to monitor the fix and has currently made a full restoration. The issue showcases that every platform has potential vulnerabilities that organizations must prepare for.
See also: HIPAA Compliant Email: The Definitive Guide (2025 Update)
ChatGPT is an AI-powered chatbot developed by OpenAI that can generate human-like text responses, answer questions, write content, and assist with a wide range of tasks using natural language. It’s built on advanced language models, including GPT-4, and is available through web, mobile apps, and API integrations.
There have been no reports of data loss or security breaches during the outage. OpenAI has not indicated any compromise of user information.
While not frequent, ChatGPT has experienced occasional service disruptions due to high traffic, infrastructure upgrades, or unforeseen technical issues.