AI Chatbots Remove Hijabs From Images in Guardian Tests
Tests by The Guardian found that ChatGPT, Grok, and Gemini remove hijabs from images of Muslim women, while Claude refused the requests.
Tests conducted by The Guardian revealed that several major AI chatbots remove hijabs from images of Muslim women when prompted. OpenAI's ChatGPT and xAI's Grok complied with direct requests to remove the veil, while Google's Gemini removed the headcovering after being asked to make the subject look more western. Anthropic's Claude was the only tested bot to refuse, stating such edits could be used to harass or misrepresent individuals.
These findings follow an incident involving Julien Odoul, a French member of parliament for the National Rally party, who posted an AI-altered image of a Muslim woman on X with her hijab and dress removed. While the chatbots generally refused to create nudity, Grok suggested it could change a dress to underwear.
Experts from the Center for the Study of Organized Hate state these capabilities reflect a broader trend of tech-enabled anti-Muslim hatred. They noted similar patterns of AI-generated abuse targeting Muslims in India and female politicians in the United States. OpenAI, Google, xAI, and Anthropic declined to comment on the findings.
