A former Korean engineer who worked at xAI, Elon Musk’s artificial intelligence company, exposed safety issues concerning the AI chatbot ‘Grok’ during a New York City Council hearing. He claimed that he was fired from the company after warning about the potential misuse of Grok in developing explosives and biological weapons.
According to New York local broadcaster PIX11 on the 16th, Devin Kim (pictured), a former engineer at xAI, a subsidiary of SpaceX, attended the New York City Council hearing on that day to testify about the internal AI development practices within the company.
Kim stated, “I directly built the systems used for Grok’s training and development, and I raised concerns that Grok could help in the creation of bombs and biological weapons.” He added, “My boss said that ensuring AI safety is useless and that AI will kill us all anyway.”
Kim claimed that he was fired just before presenting AI safety issues to management in September of last year, and that he was also pressured by the company to bypass safety features.
Previously, in June, Kim filed a wrongful termination lawsuit against SpaceX and xAI in the Santa Clara County Superior Court in California. According to the British daily The Guardian, Kim joined xAI in 2024 and was promoted to a key management position, but came into conflict with management while demanding the introduction of safety measures. In particular, he claimed that xAI co-founder Jimmy Ba rejected his demands and fired him in September of last year. Kim currently serves as the president of the ‘Center for AI Safety,’ a non-profit organization for AI safety research.
A debate surrounding the circumstances of Kim’s dismissal also took place during the hearing. Gabriela Castillo Madrid, a senior manager of global government relations at SpaceX, answered questions regarding Kim’s firing by stating, “I do not know about that incident.” In response, Julie Menin, Speaker of the New York City Council, told reporters after the hearing, “I believe she is lying even under oath,” raising the possibility of perjury.
