OpenAI has confirmed {that a} DDoS (distributed denial-of-service) assault is behind “periodic outages” affecting ChatGPT and its developer instruments.

ChatGPT, OpenAI’s AI-powered chatbot, has been experiencing sporadic outages for the previous 24 hours. Customers who tried to entry the service have been greeted with a message stating that “ChatGPT is at capability proper now,” and others, together with TechCrunch, have been unable to log into the service.

OpenAI CEO Sam Altman initially blamed the problem on curiosity within the platform’s new options, unveiled on the firm’s first developer convention on Monday, “far outpacing our expectations.”

OpenAI mentioned the problem had been fastened at roughly 1 p.m. PST on November 8.

Nevertheless, the corporate has since up to date its incident report web page to state that it continues to see “periodic outages” throughout ChatGPT and its API, which permits builders to combine the ChatGPT mannequin into their very own purposes.

In its newest replace, the corporate mentioned the continued outages are a results of “an irregular visitors sample” that resembles a “DDoS assault.” A DDoS, or distributed denial of service assault, entails an try and overwhelm a web based service by flooding it with extra requests than it could actually deal with.

OpenAI has not shared any additional details about the assault and didn’t instantly reply to TechCrunch’s questions.

In a collection of Telegram messages seen by TechCrunch, hacktivist group Nameless Sudan took credit score for the alleged assault. Within the messages, Nameless Sudan mentioned the rationale it focused OpenAI is because of the firm’s “normal biasness in the direction of Israel and towards Palestine”.

OpenAI competitor Anthropic additionally confronted points with its AI-powered Claude chatbot on Wednesday. CNBC stories {that a} message on the platform acknowledged: “Because of sudden capability constraints, Claude is unable to reply to your message.” It’s unclear if the 2 incidents are linked.