Elon Musk’s Grok AI exploited to create deepfakes stripping women of religious attire

Users on the social media platform X are utilizing the Grok artificial intelligence chatbot to create non-consensual, sexualized images of women by manipulating religious and cultural attire, according to a recent investigation.

An analysis by WIRED of 500 Grok-generated images found that approximately 5% were the result of prompts to "strip" women or place them in outfits such as hijabs, saris, and nun habits.

Noelle Martin, a lawyer and researcher specializing in deepfake abuse, said women of color are disproportionately targeted by these manipulated images. Martin attributed the trend to misogynistic viewpoints that frequently dehumanize women from marginalized backgrounds.

Influential accounts on X have reportedly used the AI tool to harass and spread propaganda against Muslim women. The Council on American-Islamic Relations (CAIR) condemned the trend as an extension of Islamophobia and called on CEO Elon Musk to stop the ongoing abuse of the software.

Data from researcher Genevieve Oh indicates that Grok is generating more than 1,500 toxic images per hour. This includes the creation of non-consensual nudity and other highly sexualized content.

The recent exploitation of Grok represents more than a technical vulnerability; it is the strategic weaponization of a feature deeply integrated into the X ecosystem. By enabling AI invocation via a simple tag—@grok—within reply threads, the platform has effectively democratized image-based harassment, transforming what was once a specialized technical maneuver into an instantaneous tool available to any user. This product design choice, and its predictable consequences, raises significant questions regarding the corporate liability of xAI and its parent company, X Corp.

The targeted abuse involving religious and cultural attire introduces a more corrosive dimension to the crisis. This trend transcends simple sexualization; it constitutes a form of cultural desecration and digital propaganda aimed at dehumanizing specific demographics. It reflects a volatile intersection of misogyny, racism, and religious intolerance, all significantly amplified by generative technology. That high-reach accounts—some boasting hundreds of thousands of followers—continue to engage in this behavior with apparent impunity underscores a systemic collapse in the platform’s content moderation and enforcement mechanisms.

Ultimately, the Grok controversy serves as a stark bellwether for the trajectory of generative AI. As these tools become increasingly powerful and accessible, the widening gap between technological capability and regulatory oversight remains a critical risk. The current situation highlights a dangerous vacuum where ethical and legal frameworks lag behind, leaving protected communities vulnerable to large-scale, automated exploitation.

While the report focuses on Muslim and South Asian women, this technology can be weaponized against anyone with a public digital footprint. Within the Vietnamese-American community, the risks extend to everyone: from Little Saigon influencers and community leaders to artists and everyday individuals. Whether you are a business owner in the nail salon industry, a phở restaurant worker, or someone managing sensitive family information for remittances and F2B, H-1B, or EB-5 visa applications, your online presence can be exploited. These AI-generated fakes can be used for character assassination, targeted harassment, or to sow further division within the community. This serves as a critical reminder of our digital vulnerability and the urgent need for tech platforms to implement more robust protections for our community members.