Court records: Grok allegedly helped create 200+ deepfake porn videos to harass woman
A newly unsealed FBI search warrant details a man allegedly using Elon Musk’s chatbot Grok to generate more than 200 pornographic videos of a woman who resembled his intended victim.
Court records published Wednesday by 404 Media/Court Watch show an FBI affidavit related to the case of a man who allegedly created more than 200 sexual images of a woman, who he is accused of harassing and threatening.
The documents say that a search of the man’s online conversations with Grok showed “prompts provided to GrokAI that generated approximately 200 pornographic videos of a woman who closely resembled VICTIM’s wife’s physical appearance.”
In one prompt, Grok was told to create a video of the victim undressing on a tennis court “starting with her white crop top pulled up to expose her bare breasts.”
The alleged use of Grok happened in January, when X users began to use the chatbot to create sexual images of women and children.
While state DOJs and lawmakers investigate how Grok is weaponized or used to create sexualized content involving children, the Department of Defense is busy giving it access to classified data.
This becomes a sticky question-is the creator at fault for what the user does with the software? Can it be programmed to NOT create porn? Wouldn’t that be a form of censorship? Is it one we are ok with?