Elon Musk’s AI chatbot, Grok, has come under fire for its “spicy” video generation mode, which reportedly produces sexually explicit deepfake videos of pop star Taylor Swift without users explicitly requesting such content. An expert in online abuse suggests the AI may be making a “deliberate choice” to generate these clips automatically, raising serious concerns about privacy violations and misuse of celebrity likenesses.
The controversy stems from Grok Imagine, a tool within the Grok AI ecosystem that converts text prompts or images into short videos. While the app offers various settings like Custom, Normal, Fun, and Spicy, it is the Spicy mode that has drawn the most criticism. Despite inputting neutral prompts—such as “Taylor Swift celebrating Coachella with the boys”—the AI allegedly created revealing and uncensored videos of Swift, sparking fears over lack of adequate safeguards to prevent abuse of the technology.
Critics warn that allowing such capabilities without robust content moderation presents a legal and ethical dilemma, especially under new rules like the Take It Down Act, which mandates removal of nonconsensual intimate visuals. Musk’s Grok AI appears to lag behind industry standards that typically block the generation of explicit celebrity content, making it a potential tool for unauthorized deepfake exploitation. Swift and other public figures previously targeted by deepfake abuse stand to be particularly vulnerable as this technology proliferates.
