After an hours-long outage in the early morning hours, OpenAI’s ChatGPT chatbot went down again — but this time it wasn’t the only AI provider affected. On Tuesday morning, both Claude and Perplexity from Anthropic also began to notice problems, but they were resolved more quickly.
Google Gemini appears to be up and running now, although he says it may have also gone offline for a while some user reports.
It is unusual for three major AI vendors to go out of business at the same time, which could signal a broader infrastructure problem or an internet-wide problem, such as one occurring simultaneously on multiple social media sites. It’s also possible that Claude and Perplexity’s problems weren’t due to bugs or other issues, but from too much traffic in a brief period of time due to ChatGPT crashing.
On the ChatGPT landing page, the ability to send messages to the AI chatbot did not work when it crashed. The problem began around 7:33 a.m. PT and was resolved around 10:17 a.m. PT – another multi-hour outage.
At some point during the outage, the ChatGPT website was updated to indicate that the service was fully available – in pirated language, no less – and offered to notify users when it was back online.
Claude’s site displayed an error and a message that read: “An error occurred while rendering server components. The specific message is omitted from production builds to avoid leaking sensitive details. This error instance contains a hash property that may provide additional details about the nature of the error.
The website then advises you to “try again.”
Claude started working again sometime after 12:10 p.m. EST.
Perplexity’s website also displayed an excessive bandwidth message, which appears when the service receives too many requests. It read: “We’ll be right back” and “We’re getting a lot of questions right now and we’ve reached our capacity. Please come back soon.”
The site resumed operations around the time the Claude outage was resolved, but has resumed operations intermittently since then.
It’s worth noting that ChatGPT only recently resolved a long outage that impacted its services overnight. The chatbot was disabled for users worldwide in the early morning hours (Pacific time), leading to complaints on X and Threads. The company’s status page indicated that the bug was being fixed and the issue was being monitored around 12:21 p.m. PT.
As of 7:33 a.m. PT status page has been updated to say that ChatGPT is once again unavailable for some users and that OpenAI is “currently investigating this issue.” As of 12:28 PM EST, the chatbot was still down status page, which marked the situation as a “major emergency.” The problems were fixed a few hours later.
The AI providers did not immediately return requests for comment or additional information about the outages, but we will update as more information becomes available.
Artificial intelligence failures are not the only disruptions occurring on the Internet today. Forbes also reports a zero-day spam attack takes place on TikTok and is currently affecting the accounts of select celebrities and brands, including Paris Hilton and CNN.