“We are aware of recent posts made by Grok and are actively working to remove the inappropriate content,” said a statement from xAI, Musk’s AI company. “We are updating the model to prevent such responses before they are ever posted.”
This isn’t the first incident of Grok going rogue. In May, the bot randomly began bringing up the topic of “white genocide” in South Africa during unrelated chats. xAI claimed that was the result of an unauthorized modification.
Political and Ethical Fallout
Musk—who now serves as a senior advisor to President Donald Trump—has become a central figure in the current administration’s Department of Government Efficiency (DOGE). According to Forbes, he’s also contributed nearly $290 million to pro-Trump causes.
His involvement in U.S. politics and federal agencies heightens the stakes surrounding Grok’s behavior. Politico warned this week that Grok’s behavior shows just how dangerous unregulated, large-scale AI may become.
When GrokAI is prompted by politically motivated rhetoric, and is too eager to please, it can get worse.
