A ChatGPT outage disrupted access for users around the world, with reports of errors, slow responses, and trouble loading the service across web and app experiences. OpenAI also reported issues affecting ChatGPT, Sora, and its API during what it described as a partial outage.
The ChatGPT outage became visible through spikes in user reports on Down Detector, with Tom’s Guide saying reports began appearing around 2:45 a.m. ET and rose again as more regions came online. TechGig, meanwhile, reported that Down Detector reports peaked at more than 12,580 during the disruption.
What users reported during the outage
Users described a range of problems, from ChatGPT failing to load to repeated error messages when submitting prompts. TechCrunch said that when it tried using GPT-4o in ChatGPT, it received an error stating, “Too many concurrent requests.”
Tom’s Guide reported that both free and paid tiers of ChatGPT were affected, and that OpenAI’s image generator Sora also experienced problems. Tom’s Guide also noted that some users could still use the service normally, while others saw recurring errors or could not load ChatGPT at all.
TechGig reported that many users struggled to log in, and some were removed from active sessions during the outage. TechGig also said users reported seeing a blocking message that read, “Unusual activity has been detected from your device. Try again later.”
What OpenAI said about the disruption
OpenAI acknowledged the disruption and said users could see “elevated errors and latency” across affected services. Tom’s Guide quoted OpenAI’s status message saying that some users were experiencing elevated error rates and latency while the company investigated.
TechCrunch reported that OpenAI began investigating late Monday night and that the partial outage continued into Tuesday morning. TechCrunch also reported that OpenAI said it identified the issue around 5:30 a.m. PT and began working on a fix, while warning at roughly 8 a.m. PT that full recovery could take “another few hours.”
TechGig later attributed the disruption to what it called a “routing misconfiguration,” saying OpenAI confirmed that was behind the spikes in errors. Tom’s Guide, however, reported that no explanation was given for the outage, despite it lasting much longer than a previous blackout it referenced.
Status page reports differ across sources
TechCrunch said OpenAI’s status page indicated the company was dealing with a partial outage affecting ChatGPT, Sora, and the API. Tom’s Guide also described OpenAI’s status reporting as showing a partial outage while the company investigated elevated error rates and latency.
TechGig, however, reported a different picture, stating that OpenAI’s official status page showed no issues even as user complaints increased. These accounts conflict on what the status page displayed during the period of heavy user reports.
Why this outage drew attention
TechCrunch described the disruption as notably long compared with typical ChatGPT outages that “last just a few hours.” Tom’s Guide similarly characterized the incident as lasting over 10 hours before normal service was restored for most users.
TechCrunch also noted the outage occurred during a period of major OpenAI-related announcements, including Apple’s WWDC news about deeper integrations with OpenAI’s models and OpenAI’s reported business updates. TechGig framed the outage as significant for developers and organizations that rely on ChatGPT and GPT tools in day-to-day workflows, especially when access issues concentrate on the web interface.
