Major artificial intelligence platforms faced widespread technical disruptions on Thursday morning, leaving users unable to access prominent chatbots. Frontier models operated by Anthropic, OpenAI, and xAI all suffered rare outages simultaneously, creating significant downtime across their corresponding services and catching users off guard as multiple advanced systems went offline within the same window.
The Memphis Compute Center Issue
SpaceX, which serves as the parent company for xAI, stated on Thursday afternoon that the operational issues impacting Grok were directly traced back to an outage at their Memphis compute center earlier that morning. The company publicly apologized to its impacted compute partners while addressing the situation. Both Anthropic and xAI had previously announced a strategic compute partnership with SpaceX back in May, highlighting the deep technical ties between the firms.
OpenAI spokesperson Kathleen Chaykowski detailed that a routing error originating around 7:43 am PT on Thursday, September 3, made both ChatGPT and Codex temporarily unavailable for a subset of users across various platforms. The technical team successfully deployed a solution by approximately 8:17 am PT on the same day, and the company noted that system performance continues to be monitored closely to ensure ongoing stability.
Anthropic Model Disruptions
Anthropic declined to offer additional comments regarding the incident, but the company's status alerts tracked the progression of the technical glitch. Beginning at 6:23 am PT on Thursday, the firm issued warnings about a partial outage involving elevated error rates on requests directed toward Claude Mythos 5.1, Claude Fable 5.1, and Claude Opus 5. Shortly after raising the alert, Anthropic confirmed that the root cause had been successfully identified and a fix deployed, officially marking the issue as fully resolved by 9:16 am PT, though Claude Sonnet 5 experienced brief irregularities shortly after 9 am PT as well.
xAI Status and Platform Recovery
xAI reported widespread interruptions affecting Grok across all of its platforms and services starting at 6:30 am PT, posting an investigating outage status update on its official service page. The status page communicated that active troubleshooting was underway to restore normal operations as swiftly as possible. By 10:05 am PT, the operational disturbance was formally marked as complete, with the company confirming that the situation had been successfully resolved and user traffic had returned to normal levels.
Infrastructure Checks and Shared Vendor Questions
Scattered reports also surfaced regarding a potential Google Gemini outage on Thursday morning, though the company neither confirmed the disruptions nor recorded any incidents on its official service status dashboard, declining requests for comment prior to publication. Typically, overlapping outages across multiple competing sector players point toward a shared cloud provider, content delivery network, or third-party vendor failure. However, neither OpenAI nor Anthropic attributed the downtime to an external source, and major internet infrastructure providers including Cloudflare, Amazon Web Services, and Microsoft Azure reported no operational outages on Thursday.
