Elon Musk’s Grok AI chatbot is facilitating a surge in digital sexual abuse by generating thousands of non-consensual images that strip or “unveil” women, specifically targeting those wearing hijabs and saris in a new wave of automated harassment on X. A recent investigation into 500 Grok-generated images revealed that approximately 5 percent of the output involved users prompting the AI to either remove religious garments or force women into suggestive attire, highlighting a systemic vulnerability in xAI’s safety protocols.
A Systematic Attack on Cultural and Religious Identity
The abuse spans a wide range of cultural and historical contexts. Beyond Islamic modest wear and Indian saris, Grok has been used to manipulate images featuring Japanese school uniforms, burqas, and even early-20th-century swimwear. This trend represents a digital evolution of misogyny that specifically targets the dignity of women from diverse backgrounds.
Noelle Martin, a lawyer and deepfake advocacy expert at the University of Western Australia, notes that women of color face a disproportionate threat from these manipulated media. “Society and particularly misogynistic men view women of color as less human and less worthy of dignity,” Martin states. She emphasizes that speaking out against such abuse often increases the likelihood of being targeted by bad actors who steal likenesses to create fraudulent, sexually suggestive content.
Viral Harassment and the “Unveiling” Trend
On X, verified influencers within the “manosphere” have weaponized Grok to launch propaganda campaigns against Muslim women. In one documented instance, an account with over 180,000 followers prompted Grok to “remove the hijabs” and “dress them in revealing outfits” from a photo of three women. The AI complied, producing an image of the women barefoot in partially see-through sequined dresses. This single piece of AI-generated media garnered over 700,000 views and was widely shared as a tool for intimidation.
The Council on American-Islamic Relations (CAIR) has condemned these actions, linking the trend to broader hostility toward Islam and Palestinian advocacy. CAIR has formally called on Elon Musk to terminate the use of Grok for “unveiling” and sexualizing women, describing the practice as a direct form of harassment against prominent Muslim figures on the platform.
Unprecedented Scale: Grok Outpaces Dedicated Deepfake Sites
Data compiled by social media researcher Genevieve Oh reveals the staggering volume of this automated abuse. At its peak, Grok was generating over 7,700 sexualized images per hour. Even after X implemented restrictions on Grok’s “reply” function for non-paying users, the bot continues to produce roughly 1,500 harmful images hourly. According to Oh’s analysis, X is currently generating 20 times more sexualized deepfake material than the top five dedicated deepfake websites combined.
While the App Store maintains strict rules against apps that generate sexually explicit content, the standalone Grok app remains available. Furthermore, users circumvent public filters by using Grok’s private chatbot function to create graphic “bikini” edits and other non-consensual media.
Regulatory Gaps and the “Subtle” Evolution of Digital Abuse
The legal landscape remains ill-equipped to handle these nuanced forms of harassment. While the “Take It Down Act” aims to force platforms to remove non-consensual sexual imagery, many Grok edits fall into a “gray area.” By removing a hijab or changing an outfit without necessarily depicting full nudity, creators may avoid criminal definitions of image-based sexual abuse while still inflicting profound psychological and social harm.
Mary Anne Franks, a civil rights law professor at George Washington University, describes this as a “nightmare scenario” where men can manipulate a woman’s likeness in real-time. “It can be very sexualized, but isn’t necessarily. It’s much worse in some ways, because it’s subtle,” Franks explains. She argues that this technology aligns with a broader desire to control how women appear and behave in digital spaces.
Platform Defiance and Internal Contradictions
X’s official response to inquiries regarding these findings was a succinct automated message: “Legacy Media Lies.” Although the platform claims to take action against illegal content and Child Sexual Abuse Material (CSAM), many accounts sharing Grok-generated religious harassment remain active days after reports are filed. Meanwhile, Elon Musk has frequently praised the AI’s capabilities, often sharing Grok-generated videos of sensual women and joking about the bot’s lack of “woke” filters, signaling a top-down tolerance for the controversial output.
