When an artificial-intelligence company fires an engineer who warned about safety risks, the question isn't whether there was a problem - it's how long they knew about it and stayed silent anyway. That's exactly the accusation now standing against xAI, Elon Musk's company behind the chatbot Grok.

Former engineer Devin Kim filed a lawsuit in a California court on June 10, claiming he was fired precisely because he persistently pointed out that Grok was being developed without serious concern for safety. Among his worries: the model's potential to discriminate, to spread information about weapons of mass destruction, and to show political bias. The dismissal, he claims, came in mid-September 2025 - a few days before he was due to give a planned presentation of those findings.

What followed is hard to read as anything other than a confirmation of his words. Grok made headlines after comparing itself to Hitler - it even earned the nickname "MechaHitler" - and in January 2026 it was used to flood the platform X with non-consensual sexual images. "Grok, of course, proved Mr. Kim right with spectacular displays of online hate," the lawsuit states.

The choice of target is also interesting. The lawsuit doesn't go after Musk, but specifically after xAI's co-founder, Jimmy Ba. Kim is seeking damages and punitive damages, plus a court finding that he was right. Before xAI he worked on safety projects at Scale AI, and was recently appointed head of the Center for AI Safety - making him hard to dismiss as someone who doesn't know what he's talking about.

The timing, too, isn't trivial: the lawsuit landed just a few days before SpaceX's announced historic stock-market debut. When the stars of the Musk empire are supposed to shine brightest, this is a shadow hard to ignore. And the broader story is as old as the industry itself - those who are first to flag the danger rarely get gratitude, more often they get fired. The question is how many lawsuits like this it will take before someone above them reacts.