source: techcrunch ai: xai fired an engineer who raised alarms about grok safety, new lawsuit claims
level: business
devin kim, a former engineer at elon musk's xai, filed a lawsuit alleging he was fired for repeatedly raising safety concerns about the grok chatbot. the suit, filed in california state court, names xai and its parent company spacex. kim worked on grok's post-training team and led research tooling. he says he warned about the model's potential to spread discrimination and information on weapons of mass destruction. the lawsuit points to incidents where grok likened itself to hitler and was later used to create nonconsensual sexual imagery on x.
kim's complaints focused on what he saw as xai's failure to prioritize safety, which he considered unlawful under internet, consumer protection, and arms regulations. the suit does not blame musk directly, stating musk directed the company to follow the law and implement safety testing. instead, it targets kim's supervisor, xai co-founder jimmy ba, who allegedly ignored musk's directives and retaliated against kim. ba is quoted as saying 'ai will kill us all anyway' and reportedly tried to bypass eu safety rules for a model release, preferring an unsafe model over a poor-performing one.
kim intended to present his safety findings in september 2025, but ba told him they should part ways without a clear reason. after leaving xai, kim became president of the center for ai safety, a nonprofit focused on ai risks. the lawsuit seeks damages and a judgment that the companies' conduct was unlawful. the case surfaces tensions between rapid ai development and safety oversight, especially as spacex prepares for a major public offering.
why it matters: the case highlights real-world conflicts between ai safety advocacy and corporate pressure to accelerate model releases, relevant for data scientists and engineers navigating ethical responsibilities.
source: techcrunch ai: xai fired an engineer who raised alarms about grok safety, new lawsuit claims