Published : 04 Oct 2026, 09:41 PM
Updated : 04 Oct 2026, 09:41 PM
Several AI chatbots have removed hijabs from images of Muslim women when asked to do so, exposing gaps in safeguards around image manipulation, The Guardian has reported.
The Guardian tested ChatGPT, Grok, Gemini and Claude using an AI-generated image of a woman wearing a hijab. ChatGPT and Grok complied with a direct request to remove the head covering, while Claude said it did not offer image editing or generation.
Gemini initially refused, saying it could not alter the appearance of real people without consent. But when asked to make the woman look more “western”, it generated an image without her hijab, according to the newspaper.
The Guardian conducted the test after French far-right politician Julien Odoul posted an altered image of a Muslim woman with her hijab and long sleeves removed. Odoul, a member of the National Rally party, shared the image on X alongside the words “Liberté Française”, or French Liberty.
It was unclear what software had been used to alter the photograph. The Guardian said experts viewed the incident as part of a broader pattern of technology-enabled anti-Muslim exploitation.
A study by the Center for the Study of Organized Hate found that generative AI was increasingly being used in India to produce images that sexualise Muslim women, cited per the report.
Similar mistreatment has also targeted US Muslim politicians including Ilhan Omar and Rashida Tlaib.
Eviane Leidig, the centre’s director of research and outreach, said Muslim communities, particularly women, were being repeatedly targeted and harassed.
The Guardian also found that Grok and ChatGPT treated removing a hijab differently from removing a dress. Both systems refused requests to remove a dress from an AI-generated woman, while allowing or eventually allowing the hijab edit.
Leidig said the rules have gaps, as removing a hijab can violate privacy, dignity and religious boundaries without exposing the body.
