OpenAI’s ChatGPT Faces Intermittent Outages Attributed to Potential Cyberattack

OpenAI’s ChatGPT, an AI chatbot, has been grappling with intermittent outages, and the company has cited a potential cyberattack as the cause. These disruptions persisted even after the platform’s recovery from a significant outage attributed to a surge in interest for new features.

OpenAI’s service status page indicated that both ChatGPT and its API services, used by developers on the platform, were experiencing “periodic outages.” OpenAI attributed these issues to an “abnormal traffic pattern” indicative of a Distributed Denial of Service (DDoS) attack—an attempt to overwhelm an online service by inundating it with more requests than it can handle. The company acknowledged the situation and stated its commitment to mitigating the problem.

Duration of the issue

As of the time of writing, the outages have persisted for over six hours, with OpenAI actively working to resolve the situation. While users have reported some problems, the scale of this issue appears smaller than the earlier significant outage experienced on the platform.

Earlier in the week, ChatGPT and its API faced a major outage that lasted nearly two hours. Users attempting to access the chatbot were met with a message indicating that “ChatGPT is at capacity right now.” OpenAI acknowledged “errors impacting all services” on its status page but was able to identify the problem and implement a fix. The platform had also encountered a partial outage the previous night.

Impact on OpenAI competitor anthropic

OpenAI’s competitor, Anthropic, also experienced difficulties with its Claude chatbot on the same day. Users seeking to access the platform were informed of “unexpected capacity constraints.” It remains unclear whether Anthropic’s outage was connected to OpenAI’s issues and whether these “capacity constraints” resulted from ChatGPT users seeking an alternative during its outage.

OpenAI CEO Sam Altman’s response

OpenAI CEO Sam Altman attributed the initial outage to an overwhelming level of interest in the platform’s new features, which were introduced on Monday. In a tweet, Altman explained that the “usage of our new features from devday is far outpacing our expectations.” He further noted that the company had planned to make GPTs available to all subscribers on Monday but had not yet been able to do so. Altman expressed hope for a resolution soon, acknowledging potential service instability in the short term due to high demand.

OpenAI’s ChatGPT, while offering powerful AI capabilities, has encountered technical challenges in managing surges of user interest and potential cyberattacks. As AI continues to advance, ensuring the stability and security of these platforms becomes increasingly vital for both developers and users. OpenAI and its competitors must navigate these issues to provide reliable and accessible AI services to the public.