- The OpenAI Artificial Intelligence model ChatGPT experienced an outage on November 8th.
- This outage was caused by a DDoS (Distributed Denial of Service) attack.
- No user information was compromised in the incident.
According to OpenAI, the recent interruption in the ChatGPT chatbot’s services may have been caused by a targeted attack on its servers.
On November 8th, users experienced intermittent disruptions while accessing OpenAI’s tools and services, with messages indicating that the platform had reached its capacity limit. Subsequently, OpenAI issued an announcement on its official website, stating that it was “addressing periodic outages caused by anomalous traffic patterns indicative of a DDoS attack.”
OpenAI CEO Sam Altman initially attributed this issue to the high interest and frequent access of users to the platform’s new features, stating that it had “far exceeded our expectations.” He also mentioned that the problem was expected to be resolved around 1 PM.
However, in the subsequent updates on the company’s incident report page, it continued to indicate “periodic interruptions” to ChatGPT and its API. The latest update showed that the ongoing interruptions were “due to anomalous traffic patterns indicative of a DDoS attack.”
DDoS attacks, also known as Distributed Denial of Service attacks, typically refer to malicious attempts by attackers to flood internet servers using various means, disrupting normal data flow.
At present, OpenAI has not shared any further information regarding the attack.
Also read: what is going on with online privacy
According to sources, the hacktivist group Anonymous Sudan claimed responsibility for the alleged attack. Despite its name indicating ties to the African nation Sudan, security researchers speculate that this organization may have connections to Russia.
On the same day, the artificial intelligence-driven Claude chatbot from Anthropic, a competitor of OpenAI, also faced similar issues. A message on the Claude chatbot platform stated, “Due to unexpected capacity limitations, Claude is unable to respond to your messages.” It is currently unclear whether these two incidents are related.
Fortunately, OpenAI has stated that no user information was compromised in this incident.
Photo by Sora Shimazaki and Airam Dato-on