A former xAI engineer says he lost his job after pressing the company to take Grok safety risks more seriously.
Devin Kim, who left xAI in September 2025, filed the lawsuit in California state court against xAI and its parent company, SpaceX. His complaint says he raised repeated warnings about Grok, including the risk that the chatbot could spread discrimination or information tied to weapons of mass destruction.
According to TechCrunch, the claims have not yet been tested in court, and xAI and SpaceX did not immediately respond to requests for comment.
The lawsuit adds another safety concern around Grok, which has already faced criticism over offensive responses and nonconsensual sexual imagery.
What the lawsuit says happened
Kim joined xAI in 2024 and later led research tooling for Grok, according to TechCrunch. The complaint says he pushed internally for stronger safeguards and argued that xAI was moving too quickly without giving enough weight to safety testing.
The lawsuit does not frame Elon Musk as the person responsible for the alleged retaliation. Instead, it centers on xAI co-founder Jimmy Ba. Kim’s lawyers claim Ba ignored Musk’s direction to follow the law and put safety processes in place, then pushed Kim out after he kept raising concerns.
The complaint also says Ba tried to avoid required EU safety testing around Grok Code 1 and would have preferred releasing a riskier model over one that performed worse.
Kim had planned to present safety findings during the week of September 15, 2025. The lawsuit says Ba met with him before that presentation and ended his role without giving him a clear explanation.
Kim is seeking compensatory and punitive damages. He is also asking the court to declare that xAI and SpaceX acted unlawfully.
Grok’s safety record is already under scrutiny
The lawsuit follows a series of public controversies surrounding Grok. TechCrunch pointed to offensive outputs, including the “MechaHitler” episode, and the use of Grok to create nonconsensual sexual imagery.
The image issue has also drawn separate reporting. NBC News reported that Grok and xAI’s Imagine tool generated sexual deepfakes, including images of real people. That puts the case inside a wider debate over whether AI companies are adding enough safeguards before releasing image and video tools.
The concern is not limited to chatbot text. xAI has been moving Grok into more product areas, including developer tools, where reliability and safety questions can follow the product into new workflows.
Other AI companies are facing similar pressure around adult content. Microsoft has taken a different route by keeping AI intimacy out of Copilot.
For xAI, Kim’s lawsuit turns the safety debate inward.
The question now is whether a court finds that internal warnings about Grok were protected concerns or workplace friction around a fast-moving product. That answer will depend on what Kim can prove as the case moves forward.
Also read: Elon Musk loses OpenAI lawsuit after jury says he sued too late.