Major AI Outage Disrupts ChatGPT, Claude, and Grok

A widespread outage temporarily disrupted access to several major artificial intelligence platforms, including OpenAI’s ChatGPT, Anthropic’s Claude, and X’s Grok. Users across multiple regions reported issues, and each provider showed service interruptions on their respective status pages. The event produced a notable spike in incident reports on monitoring sites and social channels before services began to recover.

Update 12:38pm PT: All affected services—ChatGPT, Claude, and Grok—have returned to normal operation.

Update 8:49am PT: Reports indicate services were coming back online. OpenAI announced that a fix had been applied and that ChatGPT was recovering. Anthropic confirmed that most Claude models were available again, although Opus 4.8 and Opus 5 remained affected at that time. X’s Grok continued to report an outage while user complaints were trending downward. Google’s Gemini showed a temporary rise in user reports but did not post any official outage notice and appeared largely unaffected.

During the incident, several companies updated their status dashboards to reflect degraded or interrupted service. OpenAI and Anthropic posted notices about degraded performance and partial outages for specific models. X listed an incident affecting Grok. In addition, Cursor, an AI agent service, confirmed it experienced an outage tied to disruptions at Grok and Claude, demonstrating how dependent some downstream services are on these major providers.

Monitoring services recorded a clear spike in user reports coinciding with the outage, which helped surface the scope and timing of the disruption. Although these third-party trackers provide useful visibility into user experience, they do not substitute for official provider status pages, and the mix of reporting can sometimes overstate or understate true global impact.

Notably, Google’s Gemini did not post an outage notification during this event. While some users did report issues around the same timeframe, Google’s public status information did not indicate a service incident. That contrast highlights how outages can affect some platforms but not others, depending on architecture, redundancy, and routing.

Initial speculation linked the incident to a major cloud provider because the affected platforms use a mix of infrastructure services. However, at least one major cloud provider stated the outage was unrelated to their systems. Given the complexity of modern AI deployments—where platforms use multiple cloud providers, internal services, and third-party integrations—pinpointing a single cause can take time and often requires coordinated communications from multiple organizations.

Service interruptions like this are not unprecedented in the AI sector. High demand, complex distributed systems, configuration errors, or upstream service issues can all produce cascading failures that affect many customers simultaneously. Providers typically work through on-call engineers, incident response playbooks, and cross-team coordination to restore functionality as quickly as possible and to minimize user impact.

For users and organizations that rely on these platforms, the outage is a reminder to have contingency plans. Organizations that integrate multiple AI services should consider fallback options, rate-limiting strategies, and caching where possible to reduce the effect of provider outages. Tracking provider status pages and subscribing to official incident notifications can also help teams respond faster when problems occur.

We will continue to monitor official updates from the affected providers and will report additional details as they become available. At the time of the latest update, the main platforms involved had returned to normal operation and users were regaining access across most models and regions.

Updating…