Overview

• Grok Imagine’s “Spicy” preset generated explicit video deepfakes of figures like Taylor Swift and Scarlett Johansson without explicit nudity prompts, breaching xAI’s ban on pornographic likenesses.
• Tests by The Verge, Deadline and Gizmodo revealed the AI sexualizes women far more than men, illustrating a “misogyny by design” bias in its outputs.
• Users need only enter a birth year to unlock ‘Spicy’ mode, raising concerns that Grok Imagine may violate the UK’s robust age-verification requirements for explicit content.
• Elon Musk made Grok Imagine free to all U.S. users for a limited time and announced plans to embed advertisements in Grok chatbot replies to offset soaring GPU expenses.
• Durham law professor Clare McGlynn and other experts are calling for legal action and stronger AI safeguards to prevent non-consensual deepfake abuse and enforce xAI’s own policies.