News
OpenAI Security system leader Heidecker is about to leave his job: taking stock of 8 core members of the team who have left in recent years
2 min read
Source: ithome.com
IT House reported on July 14 that according to a report by Business Insider on the 13th, Johannes Heidecker, head of the OpenAI security system team, is about to leave the company. The OpenAI Security and Alignment team has seen several employees leave in recent years. Chen Xinhan, chief research officer of OpenAI, told Business Insider: "We are grateful to Johannes for his contribution to OpenAI. Research and security work will enter a new stage under the leadership of Mia Glazer, and we are very excited about it." OpenAI said that Glazer will serve as vice president of research and security and promote security work to be more deeply integrated into various research teams. This adjustment shows that OpenAI is rethinking the relationship between research and security work. A spokesman for OpenAI said: "Only by understanding the basic capabilities of the model can we make correct security decisions, and only by understanding the potential security impacts can we make correct research decisions." OpenAI said that it is OpenAI's mission to ensure that AGI benefits all mankind when it is eventually released. "The so-called security is to release positive value while reducing the negative impact of AI. Therefore, security is the core of our mission." However, OpenAI has been unable to stably retain senior security leaders, and some departing employees have publicly questioned whether OpenAI truly values security work. Jan Lake, who co-led the OpenAI super-alignment team, bluntly stated when he left in May 2024 that OpenAI was more focused on launching products than ensuring product safety. "In the past few years, safety culture and safety processes have been pushed to the back burner by glossy products." Lake also said that OpenAI must invest more resources to prepare for more powerful models in the future, and building machines that are smarter than humans is an inherently dangerous enterprise. Well-known AI analyst Andrew Curran borrowed a phrase from "Harry Potter" and described the phenomenon of OpenAI's rapid turnover of security leaders as a curse. The following are the departures of O in recent years