OpenAI, Claude, Grok and Gemini Hit by Rare Overlapping AI Outages
OpenAI, Anthropic, xAI and Google AI services suffered rare overlapping outages, disrupting Claude, ChatGPT, Codex, Grok and Gemini for hours Thursday.
Summary
On Thursday, September 3, 2026, cloud AI services from Anthropic, OpenAI, xAI and Google suffered rare overlapping interruptions. Anthropic reported elevated errors at 9:23 a.m. Eastern affecting Claude Mythos 5.1, Claude Fable 5.1 and Claude Opus 5, identified the cause roughly 15 minutes later, deployed a fix and resolved the partial outage by 12:16 p.m. Claude Sonnet 5 separately experienced brief elevated errors just after noon. OpenAI reported degraded ChatGPT and Codex performance at 10:43 a.m., applied mitigation a little over 30 minutes later and declared the incident resolved at 12:55 p.m.
Grok remained impaired at publication as xAI worked to restore service. DownDetector reports rose from fewer than 10 just before 9 a.m. to 1,365 by 9:45 a.m., then fell to 273. Google did not publicly acknowledge a Gemini incident, although reports climbed from 23 around 10:30 a.m. to 412 just after 11 a.m. StatusGator identified a likely Gemini API outage from 10:45 to 11:15 a.m. before service normalized. Amazon Web Services, Microsoft Azure and Cloudflare reported no major incidents despite smaller DownDetector spikes.
Interruptions across all four frontier AI providers within hours are practically unheard of. Over the previous 90 days, Claude reported 99.4 percent uptime, ChatGPT 99.63 percent and ChatGPT Codex 100 percent. Claude last recorded a similar three-hour partial outage on August 24, while ChatGPT Work Mode experienced hours of elevated latency on August 31.
Positives
- Anthropic identified the cause within roughly 15 minutes, deployed a fix and restored affected Claude models by 12:16 p.m.
- OpenAI mitigated degraded ChatGPT and Codex performance within about 30 minutes and closed the incident at 12:55 p.m.
- Grok reports fell from a peak of 1,365 to 273 by publication.
- StatusGator showed the Gemini API returning to normal after a likely 30-minute outage.
- Amazon Web Services, Microsoft Azure and Cloudflare reported no major incidents despite smaller complaint spikes.
Risks & concerns
- Four major frontier AI providers experienced overlapping interruptions within hours, an occurrence described as practically unheard of.
- Grok remained impaired at publication, with xAI still working to restore service.
- Anthropic errors affected Claude Mythos 5.1, Fable 5.1 and Opus 5, followed by a separate Claude Sonnet 5 incident.
- Gemini complaints reached 412 while Google had not publicly acknowledged a service problem.
- Claude had suffered a similar three-hour outage on August 24, and ChatGPT Work Mode experienced elevated latency on August 31.