Over the past day or so, ChatGPT users noticed a number of outages for the AI chatbot that were likely caused by a distributed denial of service (DDoS) attack.
CNBC reported that the first outage began at about 9 a.m. on November 9. About an hour later, OpenAI posted information on its status page that it had fixed the problem.
However, the status page later confirmed multiple outages that affected ChatGPT and the API (application program interface). services The status page says OpenAI’s API and ChatGPT have been experiencing “slow response.”
The last message on the OpenAI site, posted last night, read as follows: We are dealing with periodic outages due to unusual traffic patterns reflecting a DDoS attack. distributed denial of service)”.
OpenAI said it is continuing to work to mitigate this situation. CNBC also reported that another AI chatbot, Anthropic’s Claude 2, also stopped working yesterday. It is unclear whether the two events are related. The DDoS attack may have occurred shortly after OpenAI held its first developers conference on September 8.
Elsewhere, the company announced a new AI product, GPTs. This chatbot allows developers and companies to customize their own version of ChatGPT to serve specific needs and services.
The company also announced a preview of the GPT-4 Turbo model, which will have a context window that will allow it to accept a text prompt equivalent to 300 pages and work faster than the current GPT-4 chatbot while reducing operating costs by up to two times.