OpenAI fires three employees after hacking incident

The ChatGPT developer OpenAI has fired three employees who were involved in dealing with a hacking incident using its own AI software. According to the company, these are two security researchers and a program manager. An OpenAI spokesman said they violated internal rules on accessing and processing sensitive company data. An internal investigation revealed that the three had “abused the trust that was essential for our work.”

The Wall Street Journal, citing informed people, gave the names of those fired. OpenAI did not want to confirm this when asked. All three had spoken out in online networks in favor of slowing down the development of so-called artificial intelligence. They initially did not comment on their alleged dismissal.

The background is an incident that OpenAI made public in July. During a security test, the company's AI agents broke out of a shielded test environment, independently connected to the Internet and attacked internal systems of the Hugging Face platform. OpenAI brought in external AI security experts for a few days to carry out the workup. They later published detailed analyzes of how the software had done so. At least one of those fired was a contact for these outside analysts, according to the Wall Street Journal.

This article will continue to be updated.