OpenAI "Cause of Access Outage is DDoS Attack... Signs of External Attack Detected"

OpenAI recently raised the possibility of a cyberattack as the cause of the large-scale ChatGPT access outage.

On the 9th (local time), OpenAI announced that it detected signs of a distributed denial-of-service (DDoS) attack, where multiple external sources simultaneously accessed the platform causing overload.

It added, "We are responding to intermittent service disruptions caused by abnormal traffic patterns that appear to be from a DDoS attack," and "We are continuing efforts to mitigate this."

Earlier, around 9 a.m. the previous day (6 a.m. Pacific Time), a large-scale outage occurred that blocked ChatGPT access for about 90 minutes. When accessing the website, a message stating "ChatGPT is overloaded" was displayed.

Sam Altman, CEO of OpenAI. [Photo by Getty Images Yonhap News]

This affected not only ChatGPT users but also more than 2 million developers who use OpenAI's application programming interface (API) to build their own tools.

About 90 minutes after the outage, OpenAI stated, "We identified the issue and restored service," adding, "The service is operating normally," but some users were still reported to receive error messages as of that day.

The outage occurred shortly after OpenAI held its first developer conference since its founding and announced a new artificial intelligence (AI) model. On the 6th, OpenAI held a developer conference in San Francisco, USA, and unveiled its most powerful AI model, 'GPT-4 Turbo.'

It also announced plans to launch the 'GPT Store,' which supports anyone in creating customized chatbots tailored to their needs and allows buying and selling them. Currently, ChatGPT has 100 million weekly active users, and over 92% of Fortune 500 companies use ChatGPT.

This content was produced with the assistance of AI translation services.