Recently, OpenAI has become the subject of public controversy due to a dispute over the departure of former employees. In response to various speculation and public doubts regarding the reasons for the three researchers' departure, the OpenAI research director has officially released a detailed official statement that clarifies all the key details of the incident.
In the statement, the official clearly stated that the termination of the employment of three employees—Jasmine, Mikita, and Tomek—last week was not due to them expressing public safety concerns or criticizing the company. OpenAI emphasized that internal discussions on intense and critical safety and research topics are highly welcomed, and it regards them as key to making correct decisions. The fundamental reason for the dismissal is that, after thorough investigation, these three individuals seriously violated the clear policy regarding the handling of sensitive information. Internal investigations also confirmed that their actions constituted a “major trust violation” that exceeds what was mentioned in the public letter. Based on this, the company has decided not to continue hiring them.
In addition to responding to personnel disputes, the statement also revealed OpenAI’s latest progress in external security assessments. The company is currently actively finalizing contracts with third-party security assessment agencies, and detailed information is expected to be officially released within the next few weeks. Introducing an independent external assessment agency and making close cooperation with external organizations a core aspect have been key priorities for the company throughout its efforts.
In addition, OpenAI acknowledged the industry consensus in its statement. Regarding the point mentioned in the former employee’s letter that “maintaining the monitorability of advanced models requires a shared commitment from the entire industry,” OpenAI expressed strong agreement and noted that monitorability has been one of its core research initiatives for a long time. In the future, the company will continue to invest substantial resources in this area and work together with the entire industry to further improve AI safety and alignment systems.