3 fired OpenAI researchers send open letter saying company practices could trigger 'chilling effect'3 名遭解雇 OpenAI 研究员发公开信，称公司做法或引发“寒蝉效应”

IT House, Oct. 9 (Xinhua) -- Three former OpenAI researchers, Mikita Ballesney, Tomek Kolbak and Jasmine Wang, who were abruptly fired last week, released an open letter today (Oct. 9). The letter mentions fears that the move will discourage other employees from expressing their opinions publicly, creating a “chilling effect” on OpenAI's corporate culture. In an open letter titled “OpenAI Can't Secure AI Alone,” the trio noted that OpenAI is “not an ordinary company” and that the ability of employees to work with external agencies to address security issues with confidence is an important part of OpenAI's working culture. The open letter was posted on social media by Barlesnee, who had previously been engaged in AI alignment research, to the safety committees of OpenAI. The other two signatories, Kolbak and Jasmine Wang, both previously engaged in AI security research, were also fired. According to Business Insider, an OpenAI spokesperson responded that the company had concluded after an internal investigation that the three had violated rules on access to and handling of sensitive information. The three researchers said on social media that the company only verbally informed them of the reason for the dismissal, and their previous actions were within their respective job responsibilities. Jasmine Wang was accused of accessing an email from an executive in response to which she said: “The reason given by the company for the dismissal is simply untenable. Ballesny posted that the company told him it no longer trusted him because he was“ communicating too much with third-party security agencies. ”" I have never disclosed the company's intellectual property information to the outside world. My work has been coordinated with my immediate supervisors, the management of the research department, and the board of directors. Kolbak also posted that the company had verbally informed him that the reason for his dismissal was “the way he communicated with METR.” “I want to make it clear that communicating with METR is my job." IT House Note: METR is a non-profit research organization responsible for assessing AI models and their security risks. In the open letter, the three expressed concern that other employees may not dare to express their opinions and work as before, which was an indispensable part of OpenAI's work until last week. “Sudden layoffs like ours, and the rush to notify the parties involved, have left the culture of openness that OpenAI used to be proud of impacted by the chilling effect. The three also made three recommendations to OpenAI: allow third-party security auditors to enter the company, ensure that cutting-edge AI models can still be effectively monitored, and support security researchers to maintain “open and transparent communication” with other institutions and personnel in the field of AI security. In an internal memo to employees, a Head of Research at OpenAI expressed strong support for the three recommendations. “We are very grateful for their contributions to AI security work, as well as their courage to speak up and question existing perspectives. We value their opinions, support their work, and have great trust in them. I must be clear that the dismissal decision has nothing to do with them raising safety concerns or speaking out publicly. Employees are always encouraged to express their opinions and will continue to do so. We don't fire employees for raising concerns. Kolbak also wrote on social media that he was concerned that OpenAI would use the dismissal of three people as an“ excuse ”to reduce its cooperation with METR.