Several prominent AI platforms—OpenAI's ChatGPT, Anthropic's Claude, and X's Grok—experienced widespread outages Thursday morning, with services recovering by early afternoon.

ChatGPT and its Codex tool reported elevated error rates starting Thursday morning. Claude experienced a partial outage across its chatbot, Claude Code, and API services. By 8:49 a.m. PT, OpenAI said a fix had been applied and services were recovering. Claude reported most models had recovered by then, except for Opus 4.8 and Opus 5. Grok remained down.

All three services returned to normal by 12:38 p.m. PT. Cursor, an AI-powered code editor, also confirmed an outage tied to the same infrastructure issue.

Google's Gemini did not report an official outage during the window, though user reports spiked.

The synchronized downtime points to a shared vulnerability. OpenAI, Anthropic, and xAI all rely on Microsoft Azure for cloud services, though some also use other providers including Google Cloud. That shared dependency creates a single point of failure for some of the industry's most critical infrastructure.

CJ Avilla, a technical staff member at Anthropic, attributed Claude's outage to an "infrastructure issue." xAI said Grok was "experiencing issues" and the company was working on a solution.

The incident exposes a structural problem in AI economics: the capital required to build and operate AI models at scale has concentrated compute workloads among a handful of hyperscale providers. OpenAI and Anthropic, despite raising billions in capital, have not yet invested heavily enough in proprietary GPU clusters or true multi-cloud redundancy to eliminate this risk.

For enterprise customers, the calculus is straightforward. Service continuity directly affects adoption rates and willingness to integrate AI into mission-critical workflows. A single region failure at Azure can cascade across multiple AI platforms simultaneously—exactly what happened Thursday.

Building proprietary compute or implementing genuine multi-cloud failover requires substantial upfront capital allocation, but it offers a defensible moat. As AI services embed deeper into enterprise operations, the company that solves infrastructure resilience first will have a competitive advantage worth the investment.