In recent days, Microsoft’s generative AI tool, ChatGPT, has been experiencing connectivity problems. The official message on the website indicates that their servers are operating at full capacity. However, an article published by Bloomberg suggests that the technology giant’s AI marvel is under a barrage of abnormal fake web traffic, effectively resulting in a Distributed Denial of Service attack, commonly known as a DDoS attack.
Microsoft has released a preliminary statement today, acknowledging that their computer networks are facing an overwhelming demand for service, potentially causing slow or unavailable connections for some users. Interestingly, these issues appear to affect users worldwide, regardless of their location. At times, the server reports heavy web traffic, while at other times, it simply displays a message saying, “oops… BRB.”
While Microsoft has described the outage as temporary, they have not officially attributed it to the work of a malicious actor or a cybercriminal group.
On the other hand, OpenAI, the organization responsible for the research and development of ChatGPT, has openly admitted that their subsidiary is grappling with a major outage. They are experiencing high error rates across their software and AI platform.
Interestingly, OpenAI recently held its first developer conference and introduced a preview of GPT-4 Turbo, an advanced version of the previous AI generative tool. This new model boasts significantly greater power and speed, with a capacity 100 times greater than the current ChatGPT, enabling it to achieve much more.
It’s worth noting that a source from the AI startup revealed that these issues are not entirely unexpected and have been occurring since the launch of Chat Generative Pre-Trained Transformer. Sam Altman, the CEO of OpenAI, had foreseen such challenges at the beginning of the year and had instructed his administrative teams to be prepared for potential web disruptions, with DDoS attacks being just one possible scenario. ChatGPT was introduced to the world on November 30, 2022, and the website includes a warning that the generated content may contain misleading information or be misused for malicious purposes, such as creating malware.
