Cloud AI services operated by OpenAI, Anthropic, xAI and Google experienced overlapping interruptions Thursday morning, Ars Technica reported. The timing was unusual because all four frontier-model providers showed signs of degraded service within a period of a few hours, though the available evidence did not establish a shared technical cause.
Anthropic reported a partial outage involving several Claude models at 9:23 a.m. Eastern, identified a cause about 15 minutes later and said the issue was resolved by 12:16 p.m. A separate Claude Sonnet incident appeared shortly after noon. OpenAI reported elevated errors across ChatGPT and Codex at 10:43 a.m., deployed mitigation and marked the problem resolved at 12:55 p.m.
xAI displayed a user-facing message saying Grok was experiencing problems, while Downdetector reports rose rapidly. Google did not publish a comparable Gemini incident notice, but user reports and an API-monitoring service indicated a shorter disruption. Those third-party signals establish that users encountered problems, but they do not supply the same technical detail as an operator’s incident report.
Major infrastructure providers including Amazon Web Services, Microsoft Azure and Cloudflare did not report a broad outage corresponding to the entire window. That makes a single public-cloud failure less evident from the current record, although shared dependencies can exist below the level of a platformwide incident and companies do not disclose every component of their model-serving stacks.
The providers normally report high availability: Ars cited 90-day figures of 99.4% for Anthropic, 99.63% for ChatGPT and 100% for Codex. The overlap was therefore rare, but coincidence remains possible. Post-incident reports, if released, would be required to determine whether traffic shifts, a common vendor, software changes or unrelated failures produced the same-day pattern.