OpenAI has fired three researchers for violating its confidential information handling policy by sharing information with an AI security organization.

It has been reported that OpenAI has fired three researchers for allegedly sharing highly confidential information with an external AI security organization.

Exclusive | OpenAI Fires Researchers for Allegedly Sharing Information with AI Safety Group - WSJ

OpenAI parts ways with 3 researchers it says mishandled sensitive information - CBS News
https://www.cbsnews.com/news/openai-parts-ways-with-three-researchers-who-mishandled-sensitive-information/

According to a Wall Street Journal report citing sources, three people have been fired from OpenAI: Tomek Korbak, a member of the company's security team, and Jasmine Wang and Mikita Bareshni, who were working on AI alignment .

An OpenAI spokesperson said, 'We have terminated our relationship with three individuals who violated our policies regarding access to and handling of highly confidential corporate information. Our investigation confirmed that these individuals mishandled confidential information outside of established internal procedures, violating our policies and undermining the trust essential to our operations.'

Following the hacking incident in which its AI attacked Hugging Face, OpenAI has been pressured by external security organizations to undergo security testing by third-party institutions. In connection with this incident, OpenAI allowed staff from evaluation firms METR and Redwood Research to visit its offices and conduct investigations. METR subsequently published a report based on information provided by OpenAI.

OpenAI reports analysis results after accidentally hacking Hugging Face; a grand cooperative operation begins when multiple isolated agents discover each other's presence - GIGAZINE

According to sources, Mr. Korbak, who was fired, was in charge of technical liaison with METR and Redwood Research.