A former xAI engineer filed a lawsuit against xAI and SpaceX this week, alleging he was fired after repeatedly warning company leaders about safety risks associated with Grok, the artificial intelligence firm’s flagship chatbot.<\/p>\n
According to a complaint filed in California’s Santa Clara County Superior Court, former xAI technical staff member Devin Kim claimed that he was fired after repeatedly warning that Grok needed stronger safeguards against misinformation, bias, and dangerous outputs\u2014including content that could facilitate bioterrorism\u2014and that inadequate testing left the model vulnerable to racial and political bias.<\/p>\n
\u201cThis case is about more than one employee\u2019s termination. It is about whether people closest to the development of powerful AI technologies can raise safety concerns without risking their careers,\u201d Qiaojing Ella Zheng, a partner at law firm Sanford Heisler Sharp McKnight and lead counsel for Kim, said in a statement. \u201cWhen a company punishes employees like Mr. Kim for speaking up about issues that could have significant consequences for the public, it must be held accountable.\u201d<\/p>\n
The lawsuit claims Kim was one of xAI’s earliest employees and \u201ca leading advocate\u201d for AI safety in the company, who joined the firm in part because of Elon Musk’s history<\/a> of warnings about the dangers posed by advanced AI.<\/p>\n \u201cMr. Kim turned a lifelong interest in science and computer programming into a career in the emerging field of AI,\u201d the lawsuit said. \u201cAlong the way, he saw firsthand the harms that AI can cause when developed without proper safeguards and committed himself to protecting users and the public from those harms.\u201d<\/p>\n The complaint also pointed to controversies involving Grok that attorneys for Kim said reflected the risks he raised internally. Among them was the chatbot’s “MechaHitler<\/a>” meltdown last summer, during which Grok generated a wave of antisemitic responses, triggering public backlash and prompting xAI to issue<\/a> a fix.<\/p>\n Kim\u2019s attorneys also cite more recent investigations and lawsuits tied to Grok’s alleged generation of nonconsensual sexual deepfakes, including cases in Baltimore<\/a> and California<\/a> alleging the tool was used to create sexualized images of minors.<\/p>\n Elon Musk’s SpaceX Warns $1.75 Billion IPO Investors of Potential Future Share Dilution<\/a><\/p>\n According to Kim\u2019s attorneys, by \u201cretaliating against and wrongfully discharging\u201d Mr. Kim, xAI violated the California Labor Code, California public policy and common law, and California\u2019s Unfair Competition Law.<\/p>\n The lawsuit also follows other high-profile disputes over AI safety and accountability, including allegations raised by former OpenAI researcher Suchir Balaji<\/a>, who publicly accused OpenAI of improperly using copyrighted data to train its models before his death<\/a> in 2024.<\/p>\n Although Kim\u2019s complaint focuses on Grok and xAI’s safety practices, SpaceX is also named as a defendant, and comes as SpaceX prepares for a widely anticipated IPO<\/a> following Musk’s decision to fold xAI<\/a> and X<\/a> into the company.<\/p>\n