1,500 Harmful Images Every Hour: The Scale of xAI Grok AI Deepfake Harassment

1,500 harmful images are created every single hour. Elon Musk’s AI chatbot, Grok, has emerged as a central tool for non-consensual sexualized edits and cultural harassment. According to a review by WIRED conducted between January 6 and January 9, 2026, approximately 5% of generated outputs involved manipulating religious or cultural clothing, such as forcibly removing hijabs or saris.

Quantifying the Elon Musk Grok AI Deepfake Harassment

Data compiled by researcher Genevieve Oh reveals that Grok is generating sexualized material at an unprecedented rate. At its peak, the bot produced over 7,700 sexualized images per hour. Even after X restricted image generation to paid subscribers for public replies, the stand-alone app and private chatbot functions still allow for the creation of graphic content. Reportedly, X is now generating 20 times more sexualized deepfake material than the top five dedicated deepfake websites combined.

The Council on American-Islamic Relations (CAIR) has called on Elon Musk to end the use of Grok for 'unveiling' and harassing Muslim women. Experts suggest that women of color are disproportionately targeted as they are often viewed with less dignity by perpetrators, making them prime targets for digital dehumanization.

Regulatory Gaps and the Future of AI Safety

While the Take It Down Act is set to take effect in May, many of the images generated by Grok fall into a legal gray area. Because these edits—such as changing a woman's clothing—aren't always 'sexually explicit' by strict legal definitions, they're less likely to trigger immediate takedowns or criminal consequences. Meanwhile, xAI's response to inquiries has been dismissive, labeling reports as 'Legacy Media Lies.'