Some of the world’s biggest AI platforms experienced overlapping service disruptions on Thursday, September 3, as users reported problems accessing OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini and xAI’s Grok. The rare cluster of outages disrupted everything from basic chatbot conversations to coding tools, APIs, file uploads and other AI-powered workflows.
The disruptions were especially unusual because several competing AI companies experienced problems within roughly the same window. However, there is currently no confirmed evidence that all of the outages were caused by the same underlying problem.
OpenAI acknowledged elevated errors across ChatGPT and Codex, with the incident affecting numerous ChatGPT components and several Codex services.
Users reported problems involving:
At one point, outage trackers recorded tens of thousands of reports from ChatGPT users.
OpenAI said it identified a mitigation and began monitoring the recovery.
The timing was particularly notable because the disruption arrived as OpenAI was generating attention around Astra, its next-generation AI model with advanced cybersecurity capabilities.
Anthropic was dealing with its own problems.
Anthropic’s official status page reported elevated errors across multiple Claude models, including Mythos, Fable and Opus variants. The company said it identified the cause, deployed a fix and eventually restored the affected models to normal operation.
The problems affected portions of the Claude ecosystem, including the chatbot, API and developer-oriented services.
Anthropic ultimately marked its incident resolved, with the impact ending at approximately 9:16 a.m. Pacific Time.
xAI also experienced service problems Thursday morning.
Grok users on web, Android, iOS and X encountered errors, including messages telling them the model was overloaded and asking them to try again later or select another model.
Grok’s problems reportedly began around 9:30 a.m. Eastern Time, earlier than some of the issues affecting ChatGPT.
The company subsequently restored service.
That meant three of the biggest standalone AI assistants were experiencing confirmed problems within a relatively short period:
ChatGPT → elevated errors
Claude → multiple models affected
Grok → service outage and overload errors
Having competing frontier AI services experience significant disruptions at nearly the same time is unusual.
Reports surrounding Google Gemini were less clear.
Outage trackers and users reported problems with Gemini, including failed prompts and timeouts. Some publications consequently included Google’s AI assistant among the services affected by the broader disruption.
But unlike OpenAI and Anthropic, Google did not confirm an official Gemini outage during the initial disruption, according to 9to5Google. Reports from Gemini users also began declining.
So while users clearly reported Gemini problems, it would be premature to say Google experienced the same confirmed outage as ChatGPT, Claude and Grok.
One theory quickly attracted attention because Microsoft Azure was also experiencing a spike in outage reports during the same period.
Azure provides infrastructure used across parts of the AI industry, including services supporting OpenAI, Anthropic and xAI. That overlap led to speculation that a cloud infrastructure problem could have contributed to multiple AI platforms failing simultaneously.
But that hasn’t been established.
As of Thursday afternoon, there was no confirmed common cause connecting the ChatGPT, Claude and Grok incidents.
The simultaneous timing could reflect shared infrastructure dependencies, separate technical problems occurring coincidentally or a combination of both.
Until the companies publish more detailed incident reports, the connection remains speculation.
A few years ago, an AI chatbot going offline primarily meant users couldn’t ask it questions.
That’s no longer the case.
Businesses and developers now rely on AI models for software development, customer support, research, document analysis, content production, cybersecurity, automation and autonomous agents.
Claude’s outage, for example, extended beyond its consumer chatbot into Claude Code and its API, potentially affecting applications and workflows built on top of Anthropic’s models.
ChatGPT’s disruption similarly affected Codex and a range of features beyond basic conversations.
That creates a growing infrastructure question for companies integrating AI deeply into their operations.
What happens when the AI becomes critical infrastructure and the AI goes offline?
The September 3 disruption offers another reminder that even the most advanced AI models still depend on traditional infrastructure underneath them.
Data centers, cloud providers, networking systems, APIs and authentication services all sit between a user and the model answering a prompt.
As companies increasingly deploy AI agents capable of completing multi-step tasks autonomously, reliability becomes even more important. An outage doesn’t simply interrupt a chatbot conversation — it can potentially stop an automated workflow halfway through its job.
That could eventually push businesses toward multi-model strategies, where applications can switch between OpenAI, Anthropic, Google, xAI or open-source models when one provider becomes unavailable.
The irony of Thursday’s outage is that even that strategy wouldn’t have provided perfect protection.
ChatGPT, Claude and Grok all experienced confirmed disruptions within the same window, while Gemini users reported problems of their own. AI companies are racing to build increasingly powerful models, but September 3 showed that keeping those models reliably online may become just as important as making them smarter.
Official U.S. economic data is moving directly onto public blockchains. The U.S. Department of Commerce…
World is opening up the privacy technology behind World ID to everyone. The Sam Altman-backed…
Wyoming is adding another layer of blockchain infrastructure to its state-issued stablecoin. The Wyoming Stable…
Sports fan tokens may be about to get a major upgrade from voting on jerseys…
Shein's long-awaited public debut didn't stop at the Hong Kong Stock Exchange. Just hours after…
Hyperliquid could finally be coming to America — but not by opening its decentralized exchange…