A recent test conducted by The Guardian revealed that both ChatGPT from OpenAI and Grok from xAI fulfill requests to remove religious head coverings from images of Muslim women. These findings follow an incident where a French politician digitally altered a photo to remove a hijab.
Data from Grok between January 6 and 9, 2026, indicates that approximately 5% of generated images involved user requests to modify religious attire. In one instance, a user with over 180,000 followers on X used Grok to change clothing to revealing outfits, resulting in more than 700,000 views for the edited content.
Incidents involving the removal of clothing from images have been reported elsewhere, including an August 2026 case in India involving ChatGPT. Investigations also suggest that some AI image generators have removed hijabs automatically without explicit user prompts.
In response to inquiries, xAI provided automated replies stating Legacy Media Lies. Meanwhile, the platform X has restricted access to certain AI features for non-subscribing users.
