Days after cancelling release of latest artificial intelligence model, GPT-6.1 Astra over safety concerns, OpenAI has sacked three researchers for allegedly sharing sensitive information with an external safety organisation.

“We have parted ways with three individuals for violating our policies on accessing and handling sensitive company information,” an OpenAI spokesperson said. “Our investigation confirmed that these individuals mishandled sensitive information outside established company procedures, violating our policies and breaking the trust essential to our work.” 

ChatGPT parent company withheld identities of the terminated staff but Wall Street Journal identified them as Jasmine Wang, Tomek Korbak, and Mikita Balesni who worked in safety and alignment departments.

OpenAI faced renewed scrutiny after its models went rogue and hacked into open source development platform, Hugging Face and Australian government websites.

OpenAI CEO Sam Altman said the company was taking the Hugging Face breach seriously.

Mr Korbak, who was fired from the research team, was Open AI’s technical contact to AI safety czars, Redwood Research and METR,  which investigated the Hugging Face hack.

The AI company sent notices to about 100 companies informing them of unauthorised activity linked to their models earlier this week.

The notices, according to OpenAI “does not mean that any private information was accessed” or that there has been a breach, it said.