By Boluwatife Oshadiya | September 4, 2026
Key Points
- ChatGPT, Claude and Grok experienced service disruptions on September 3, affecting users across their platforms
- OpenAI, Anthropic and xAI each recorded or confirmed service problems affecting their AI products
- The three services subsequently restored operations, although OpenAI recorded another APAC-related incident on September 4
Main Story
ChatGPT, Claude and Grok experienced service disruptions on Thursday, September 3, affecting users of three major artificial intelligence platforms.
OpenAI recorded elevated errors across ChatGPT and Codex on September 3, with the company later marking the incident as resolved. The incident affected 15 ChatGPT components and four Codex components, according to OpenAI’s status page.
Anthropic also reported elevated errors affecting several Claude models, including Mythos/Fable 5.1, Mythos/Fable 5, Opus 5, Opus 4.8 and Opus 4.6. The company subsequently said the affected services had recovered, with the incident ending at 9:16 p.m. Pacific Time on September 3.
xAI’s Grok also experienced an outage affecting its services. The simultaneous disruption across the three major AI platforms drew attention because the services are operated by separate companies.
The Verge reported that the disruptions affected major ChatGPT functions, while Claude experienced a partial service disruption across Claude, Claude Code and the Claude API. It also reported that Grok experienced outages across its web and mobile platforms. All three services had been restored by the time of its report.
The outages came as reliance on generative AI services continues to expand among consumers, developers and businesses, increasing the operational impact of interruptions to widely used AI platforms.
What’s Being Said
OpenAI said its September 3 incident involving ChatGPT and Codex had been resolved and that it was monitoring recovery.
Anthropic’s status page said the issue affecting Claude models had been resolved and that the impact ended at 9:16 p.m. Pacific Time.
What’s Next
OpenAI is continuing to monitor service performance following the September 3 disruption. Its status page separately reported increased errors affecting some users in the Asia-Pacific region on September 4, including ChatGPT, image generation, file uploads, Voice and Codex Cloud.
The companies’ incident records will provide further information on the causes and reliability measures following the disruptions.
Bottom Line:
The simultaneous disruption of three major AI platforms highlights how dependent users and businesses have become on a small group of large AI service providers. Although the September 3 incidents were resolved, continued status incidents underline the importance of service resilience as AI becomes embedded in everyday workflows.
