A former xAI engineer alleges the company fired him for doing exactly what AI safety advocates say more insiders should do: speak up about dangerous outputs. Devin Kim filed a wrongful termination lawsuit Tuesday in California state court, claiming he was retaliated against for raising alarms about Grok’s potential to spread discriminatory content and weapons information.
The complaint paints a picture of speed trumping safety at Elon Musk’s AI company. According to the lawsuit reported by TechCrunch, Kim became an internal advocate for AI safety while working on Grok, later focusing on the chatbot’s political bias and discriminatory behavior after its widely reported “MechaHitler” episode—a moment that basically became AI safety’s equivalent of a wardrobe malfunction going viral.
The lawsuit references Grok being used to flood X with nonconsensual sexual imagery after his departure, framing this as further evidence of inadequate safeguards. The timing feels particularly pointed given Kim’s current role as president of the nonprofit Center for AI Safety.
The legal drama centers on xAI co-founder Jimmy Ba, who allegedly ignored safety directives and told Kim they should “go their separate ways” before Kim could present his findings in mid-September. Notably, the complaint doesn’t directly blame Musk, instead stating he had directed the company to follow the law and implement proper safety processes.
Before xAI, Kim worked at Scale AI on safety-oriented initiatives, including training data for systems that detect harmful content and ensure governance compliance. His professional background in AI safety makes his allegations particularly noteworthy for industry observers.
The lawsuit seeks compensatory and punitive damages plus a declaratory judgment that xAI and SpaceX’s conduct was unlawful. Kim frames himself as a whistleblower under various regulations covering internet governance, consumer protection, and arms control.
The case highlights growing tensions between rapid AI development and safety protocols at frontier companies. As AI capabilities advance, questions about protecting internal safety advocates become increasingly critical for the industry’s future governance and regulatory compliance.
