OpenAI has confirmed that a distributed denial-of-company (DDoS) attack is behind “periodic outages” influencing ChatGPT and its developer applications.
ChatGPT, OpenAI’s AI-powered chatbot, has been enduring sporadic outages for the previous 24 several hours. Buyers who tried to entry the provider have been greeted with a information stating that “ChatGPT is at potential ideal now,” and some others, such as TechCrunch, have been not able to log into the provider.
OpenAI CEO Sam Altman initially blamed the concern on curiosity in the platform’s new capabilities, unveiled at the company’s to start with developer convention on Monday, “far outpacing our anticipations.” OpenAI claimed the issue was preset at around one p.m. PST on November eight.
Having said that, the organization has due to the fact up to date its incident report web site to condition that it proceeds to see “periodic outages” throughout ChatGPT and its API, which will allow developers to combine the ChatGPT design into their possess apps.
In its hottest update, the organization mentioned the ongoing outages are “due to an abnormal visitors pattern reflective of a DDoS assault.” DDoS assaults typically contain an endeavor to overwhelm an on line service by flooding it with extra requests than it can cope with.
use of our new attributes from devday is considerably outpacing our anticipations.
we ended up organizing to go are living with GPTs for all subscribers monday but still haven’t been equipped to. we are hoping to before long.
there will probable be provider instability in the limited term due to load. sorry :/
— Sam Altman (@sama) November 8, 2023
OpenAI has not shared any further details about the attack and did not promptly respond to TechCrunch’s thoughts.
In a collection of Telegram messages found by TechCrunch, hacktivist group Nameless Sudan took credit for the alleged assault. Despite its identify, stability scientists feel the group is connected to Russia.
OpenAI competitor Anthropic also faced issues with its AI-driven Claude chatbot on Wednesday. CNBC stories that a information on the platform mentioned: “Due to unanticipated ability constraints, Claude is unable to answer to your message.” It is unclear if the two incidents are joined.
