OpenAI’s ChatGPT is hit with outages. Here’s what to know.

ChatGPT stopped functioning for many users on Thursday afternoon, with OpenAI saying that its AI app was experiencing glitches for some. The app had partially recovered after a few hours, the San Francisco-based company said.

“ChatGPT, the API and Sora are currently experiencing high error rates. The issue is caused by an upstream provider and we are currently monitoring,” the artificial intelligence company posted in an update to its status page after the technical glitches cropped up at about 1:30 p.m. Eastern.

As of Thursday evening, the company said Sora, its video generation model, was fully operational and its automated application programming interface, or API, which lets software programs speak to one another, was starting to recover.

“We are continuing to work on an overall fix for ChatGPT and APIs,” the company posted as of 6:16 p.m. Eastern. About an hour earlier, it said ChatGPT had partially recovered while chat history was still not loading.

Is ChatGPT down?

More than 15,000 incidents were reported by OpenAI users on Thursday afternoon, with most of the problems related to ChatGPT, according to Downdetector, an online platform that provides users with real-time information about the status of various websites and services. 

The number of reports had dwindled to less than 700 by early Thursday evening. 

Launched in 2022, OpenAI’s ChatGPT service can generate human-like replies based on user prompts, and as of late this summer had more than 200 million active users.

According to the company, a majority of Fortune 500 companies use OpenAI’s products and its API.

OpenAI CEO Sam Altman is among the tech leaders planning to donate $1 million to President-elect Donald Trump’s inaugural fund, a spokesperson confirmed earlier this month. 

contributed to this report.