Users reported seeing an error message and instances of latency throughout the day.
Generative AI company OpenAI has today (10 June) experienced widespread disruption to its services, including popular chatbot ChatGPT, with users from across the world reporting technical issues. Users from regions such as Australia, Egypt, the UK and the US, among others, have been experiencing disruptions since as early as 7.30am, when OpenAI first acknowledged the ongoing issues.
According to TechRadar’s live coverage, beginning this morning, OpenAI stated that “some users are experiencing elevated error rates and latency across the listed services” and that they were continuing to investigate the issue. By 11am, real-time outage monitoring website Downdetector had received more than 1,000 reports of outages.
According to numerous reports, users trying to access ChatGPT have been experiencing a number of issues, including long response times from the chatbot and error messages about network connectivity.
OpenAI explained that there are several affected components across the company’s servers, which includes 14 affected API components, 21 affected ChatGPT components and four affected Sora components.
Close to 4pm, OpenAI updated its page and said: “We are continuing to work on implementing the mitigation and we are now seeing recovery on API. Full recovery across all listed services may take another few hours. We will continue to provide updates as progress is made.”
In a post on X, the company stated it has determined the root cause and its technicians are working to rectify the problem as soon as possible.
Commenting on the outage, NordVPN’s CTO Marijus Briedis highlighted the potential data risks of an incident such as this.
“This isn’t the first time that ChatGPT has experienced an outage and it won’t be the last,” he said. “However, we know that past incidents have seen some users’ conversation history and potentially payment information exposed due to bugs.”
In March 2023, OpenAI took ChatGPT offline temporarily due to a bug in an open-source library which allowed some users to see titles from another active user’s chat history. The company also said at the time that the bug may have caused the “unintentional visibility of payment-related information of 1.2pc of the ChatGPT Plus subscribers who were active during a specific nine-hour window”.
“Even temporary disruptions can sometimes reveal sensitive data,” said Briedis.
“The outage happened right in the middle of the working day, as millions look to the chatbot to help them with their daily office tasks. However, the timing should be a stark reminder that even the most sophisticated technology can be flawed.”
ChatGPT and rival AI models are becoming increasingly popular worldwide and there is a growing increase in their use on a day-to-day basis in both professional and personal capacities. OpenAI recently announced it would be expanding in Seoul, Korea, with a new office to be based there.
The company has also experienced some recent legal woes as it appealed a court order which would force the AI provider to preserve its ChatGPT logs, citing the argument that the order is in direct conflict with OpenAI’s commitment to privacy.
Last month, a US district court judge decided that OpenAI needs to preserve and segregate all output log data. This includes the free, Plus, Pro and Team versions of ChatGPT, as well as its API.
The plaintiffs – The New York Times, The New York Daily News and the Centre for Investigative Reporting – have argued for the preservation of information, stating that this data is necessary to show that OpenAI’s models copy, paraphrase and misattribute work in its output.
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
