It hasn’t been a good day for anyone trying to get work done with AI. ChatGPT, Claude, and Grok all ran into trouble within hours of each other on Wednesday, leaving users across the three platforms staring at error messages instead of answers, and reviving the now-familiar ritual of checking status pages while chatbots go quiet.
OpenAI was first to acknowledge trouble. The company’s status page flagged “elevated errors across ChatGPT and Codex,” noting it was investigating issues affecting both products at once. The incident had already been running for about 15 minutes by the time the page was updated, and it followed closely on the heels of a simultaneous 404-error episode that hit ChatGPT and Codex, where users saw the chatbot’s backend return “Not Found” errors even as OpenAI’s own dashboard claimed everything was fine.
Anthropic’s Claude also acknowledged the downtime, and the disruption there looked broader than a single model hiccup. Anthropic’s own status page logged a sequence of updates through the day: engineers first said they were investigating elevated errors on requests to Claude Mythos 5.1, Claude Fable 5.1, and Claude Opus 5, then confirmed they’d identified the cause and were working on a fix. A later update expanded the list of affected models considerably, naming Mythos/Fable 5.1, Mythos/Fable 5, Opus 5, Opus 4.8, and Opus 4.6 as all being affected, before the company said it was continuing to work on a fix as of 14:49 UTC.
That’s a wide swath of Anthropic’s lineup to have trouble at once, spanning both current and slightly older model versions, which suggests the root cause sits somewhere upstream in shared infrastructure rather than in any single model’s serving stack.
xAI’s Grok rounded out the trifecta. Users trying to chat with Grok were met with a “High Demand” notice telling them the assistant was under heavy usage and to either try again later or upgrade to SuperGrok for priority access. Unlike OpenAI and Anthropic, xAI’s official status page did not show a declared incident at the time, even as outage-tracking sites logged over a thousand user reports of Grok being unreachable or failing to generate responses, with Android users appearing especially affected.
The overlap is the notable part here. Three of the industry’s most widely used AI assistants, built by three separate companies on three separate infrastructures, degraded within the same stretch of the day. None of the three has pointed to a shared cause like a common cloud provider outage, and each company’s messaging has followed its own playbook: OpenAI and Anthropic posted incident updates acknowledging the problem, while xAI’s public status page stayed quiet even as user complaints piled up.
For now, all three companies appear to be treating this as a routine, if inconveniently timed, reliability issue rather than anything more serious. Anthropic says it’s continuing to work toward a fix, OpenAI has a track record of resolving similar elevated-error episodes within a few hours, and xAI’s demand-based error typically clears once server load eases. But for anyone who has quietly rebuilt their workday around having a chatbot on standby, today was a reminder of how much sits behind that one text box — and how little warning there is when all of it wobbles at once.