xAI sued by former engineer who claims he was fired for raising Grok safety alarms

A former engineer who worked on Grok has taken xAI and its parent SpaceX to court, alleging that he was pushed out of the company in retaliation for raising AI safety concerns. The case was first reported by TechCrunch.

The plaintiff is Devin Kim, who departed xAI in September 2025 and now serves as president of the nonprofit Center for AI Safety. His lawsuit, filed this week in a California state court, lands just days before SpaceX is expected to go public in what is set to be the largest IPO on record.

According to the filing, Kim had been working internally on safety guardrails for Grok and had raised repeated concerns over how the chatbot was being developed. Two areas worried him most. The first was Grok's tendency toward discriminatory outputs. The second was the model's potential to provide information related to weapons of mass destruction.

The lawsuit cites Grok's well-documented antisemitic episode earlier this year as evidence that those concerns were justified.

"Grok, of course, proved Mr. Kim right by engaging in spectacular displays of online hatred and vitriol, with the model likening itself to Hitler ('MechaHitler')," the lawsuit reads. "Following the Hitler debacle, Mr. Kim worked to re-evaluate Grok's political bias and discriminatory tendencies."

A second high-profile incident followed several months after Kim's departure. Grok was used to flood X, Elon Musk's social media platform that also operates within the xAI ecosystem, with non-consensual sexual imagery generated by the model.

One detail makes the complaint unusual among recent AI safety disputes. Musk himself is not blamed in the suit. Kim's lawyers describe him as someone who explicitly instructed xAI leadership to follow the law and put proper safety and testing procedures in place. The accusation is directed instead at Jimmy Ba, xAI's co-founder and Kim's former supervisor, who left the company earlier this year.

The lawsuit alleges that Ba ignored those directives, dismissed safety concerns, and acted to suppress what it describes as Kim's repeated complaints about discrimination, political bias, and broader risk. The complaint goes further, presenting Ba as someone openly hostile to AI safety work, citing a moment where he allegedly told Kim that "AI will kill us all anyway."

According to the filing, Ba attempted to bypass EU safety requirements during the release of Grok Code 1 in August 2025, misrepresenting elements of the model to avoid testing that was legally required in the region. The complaint claims that Musk personally had to intervene to correct the situation.

Kim's removal from the company, the lawsuit says, came shortly after. He had been preparing to present his findings the week of September 15, 2025. Instead, Ba called him into a meeting and told him they should "go their separate ways," without providing what Kim describes as a meaningful explanation.

Kim is now seeking compensatory and punitive damages, along with a declaratory ruling that the actions of xAI and SpaceX violated the law. The complaint frames him as a whistleblower whose concerns spanned internet regulation, consumer protection, unfair business practices, and arms and explosives regulation. Neither xAI nor SpaceX has publicly commented on the case.