SpaceXAI confirms Memphis data center caused Grok, Claude outages

Three of the world's most widely used AI platforms — OpenAI's ChatGPT, Anthropic's Claude and xAI's Grok — went offline within minutes of each other on September 3, disrupting millions of users and exposing the fragile infrastructure underpinning the AI industry. Services were fully restored by early afternoon Pacific time, but the episode has prompted urgent questions about shared dependencies among ostensibly competing platforms.

Two Failures, Not One

Despite the near-simultaneous timing, emerging evidence points to two distinct causes rather than a single cascading failure. Claude and Grok opened incident reports just four minutes apart — at 1:26 PM and 1:30 PM UTC, respectively — while ChatGPT's problems surfaced 92 minutes later, at 2:58 PM UTC.cyberkendra

SpaceXAI confirmed that Grok's outage originated at its Memphis compute center and issued an unusual apology to "impacted compute partners," the first acknowledgment by any of the three labs that outside customers were affected. Anthropic, which announced a compute partnership with xAI's parent in May 2026 and leases capacity in Memphis, has not publicly confirmed the Memphis link, though the four-minute gap between its outage and Grok's strongly suggests a shared facility was involved.forbes+1

ChatGPT's failure, by contrast, appears unrelated. An OpenAI engineer who identified themselves as the incident commander wrote on Hacker News that "a routing error within our infra" caused the disruption, adding that it was not connected to the GPT-6 Astra launch that followed on September 4. OpenAI's status page flagged "elevated errors" across ChatGPT and Codex. Downdetector logged more than 37,000 user reports for ChatGPT, along with roughly 1,300 each for Claude and Grok.thedailystar+2

Microsoft ↘ 0.99% logged no Azure incident during the window, and Cloudflare ↗ 9.91% said its infrastructure was operating normally, effectively ruling out two widely circulated theories.techround+1

Hidden Dependencies, Shared Risk

The episode laid bare a structural reality that most users never see: rival AI companies often run on overlapping infrastructure. As Jeff Watkins, chief AI officer at NorthStar Intelligence, put it, "We tend to think of OpenAI, Anthropic, Google Alphabet Inc. ↗ 0.78% and xAI as completely separate competing platforms. Underneath them, however, the AI ecosystem is much more interconnected."techround

Downstream tools felt the impact immediately. Cursor, a coding environment used by millions of developers, confirmed its own outage because it depends on Claude and Grok under the hood. Sayali Patil, CEO of IntentOps, raised a concern few others addressed: "An agent isn't a person typing a question — it's mid-execution. When the underlying model provider vanishes mid-action, that work doesn't gracefully pause. It just stops."forbes+1

The Infrastructure Question

The five-hour disruption has intensified debate over whether enterprises are adequately prepared for AI downtime. Rob Demain, CEO of e2e-assure, framed the core challenge: "If critical functions run on cloud-based AI — whether that is your SOC, your service desk or your production line — can you still run the business when it goes down? For most organisations, that answer is no, and nobody has tested it."techround

As of Friday, neither OpenAI nor Anthropic had published a full technical postmortem.cyberkendra