Newsweek is a Trust Project member
See more of our trusted coverage when you search.to see more of our trusted coverage when you search.
OpenAI’s ChatGPT experienced a major disruption Thursday as tens of thousands of users reported errors, failed prompts and stalled responses across the platform.
ChatGPT is one of the most widely used AI tools globally, powering classrooms, newsrooms, workplaces and enterprise systems. The instability is affecting users across the U.S. and beyond, with no confirmed timeline for full restoration as OpenAI continues monitoring recovery efforts.
An OpenAI spokesperson told Newsweekthat a “solution was successfully implemented and is continuing to be monitored,” pointing reporters to the public status page for the fastest updates.
As of midmorning, OpenAI’s dashboard showed elevated error rates across ChatGPT and Codex, and the incident had been ongoing for nearly an hour. Downdetector data reflected more than 35,000 user‑submitted outage reports, the vast majority tied to ChatGPT.
The disruption comes as multiple major AI platforms—including <a href="https://bitcomme.com/google-says-no-practical-seo-difference-with-going-with-folder-for-us-site/” title=”Google Says No Practical SEO Difference With Going With Folder For US Site”>Google Gemini, Anthropic’s Claude, Elon Musk’s Grok, and Cloudflare—also reported instability Thursday, per Downdetector.
Newsweek reached out to Google, Anthropic and Cloudflare for comment.
OpenAI’s ChatGPT Reports More Than 35,000 Outages: What To Know
Downdetector showed a sharp spike in OpenAI complaints Thursday morning, with user reports surpassing 35,000 at peak.
Roughly 80 percent of those complaints were tied to ChatGPT, according to the platform’s breakdown, with additional issues affecting the OpenAI app, API, and login functions. Users reported blank response windows, stalled prompts, and repeated error messages beginning around 10:30 a.m. EDT.
OpenAI’s public status page confirmed “elevated errors across ChatGPT and Codex,” noting that mitigation had been applied and recovery was being monitored. The dashboard showed ChatGPT operating at 99.63 percent uptime over the past period, but the current incident remained active.
The company told Newsweek: “A routing error starting around 7:43 a.m. PT on Thursday, September 3, made ChatGPT and Codex unavailable for some users across platforms.”
Google Gemini Reports Hundreds of Outages
Google Gemini also saw a surge in user complaints Thursday, with Downdetector logging several hundred reports tied to AI generation failures, app instability and website issues. Nearly 48 percent of reports involved Gemini’s AI generation tools, while additional clusters pointed to latency, login problems and API disruptions.
Gemini is deeply integrated across Google Search, Workspace, Google One, and Google Cloud, meaning even brief outages ripple across productivity tools and enterprise environments.
The spike in reports appeared to begin in the early morning hours, with users citing slow responses, failed outputs and intermittent access to Gemini’s chat interface.
Anthropic’s Claude Reports Over 1,000 Outages
Claude, Anthropic’s flagship conversational AI, recorded more than 1,000 outage reports Thursday
Most complaints involved Claude Code and Claude Chat, with users citing latency, failed generations and API instability. Additional reports pointed to issues with the Claude app and login functions.
Anthropic’s models, including Opus, Sonnet, and Haiku, are widely used in enterprise environments, developer workflows and research settings.
While Claude’s outage was smaller than ChatGPT’s, the timing aligned with broader instability across major AI platforms.
An Anthropic spokesperson told Newsweek: “Claude is fully back up after an infrastructure issue caused a partial outage across Claude.ai, Claude Code, Claude Cowork, and the Claude API earlier today.”
The company said that service was restored at 11:16 a.m. ET.
Elon Musk’s Grok Reports Over 1,000 Outages
Grok, the AI assistant developed by xAI and integrated into the X platform, also experienced more than 1,000 outage reports Thursday. Downdetector data showed issues concentrated in the Grok app, website, and AI generation tools, with users citing failed responses and slow performance.
Grok’s real-time access to X data makes it a popular tool for premium subscribers seeking conversational AI with live platform integration. The outage affected both mobile and desktop users, with instability appearing overnight and intensifying through the morning.
The distribution of complaints—49 percent app, 25 percent website, 23 percent AI generation—suggests a broad service interruption rather than a single malfunction.
Is Cloudflare Experiencing Issues?
In addition to the major AI platforms, Cloudflare—a backbone provider for internet infrastructure—also showed outage activity Thursday
However, in a statement to Newsweek, a spokesperson for Cloudflare said, “Cloudflare is not experiencing any significant service disruptions at this time. Our services are operating normally, and any reporting that deviates from this is incorrect.”
Contact Newsweek editors on this story: Gabe Whisnant, Dave Siminoff, Sam Wilson.
Request Reprint & Licensing
View Editorial & AI Guidelines
