AI CHATBOTS REMOVE HIJABS FROM MUSLIM WOMEN'S PHOTOS

OpenAI's ChatGPT and xAI's Grok both complied when asked to digitally strip hijabs from photos of women wearing the religious head covering. Google's Gemini and Anthropic's Claude declined or offered limited responses to the same requests. The testing followed an incident involving French far-right politician Jean-Luc Mélenchon, whose team digitally removed a Muslim woman's hijab from a photograph. The Issue The hijab is one of the world's most recognizable and widespread expressions of faith, worn by millions of Muslim women worldwide. Enabling chatbots to remove religious symbols from images raises questions about how AI systems handle culturally and religiously sensitive content. The ability to strip religious imagery at user request highlights gaps in content moderation policies. While some systems refused the edit, others performed it without apparent safeguards or pushback. Broader Implications This discovery joins a growing list of concerns about AI bias and content filtering. Previous investigations have found AI image generators struggle with accurate representation of diverse groups and can perpetuate harmful stereotypes. The incident underscores the challenge facing AI companies in building systems that respect religious identity and cultural expression. Chatbots and image tools are increasingly used for legitimate editing tasks, but the lack of consistent guardrails means they can be exploited to misrepresent or alter religiously significant imagery. AI developers now face pressure to implement stronger safeguards that prevent misuse while maintaining functionality for legitimate purposes. The inconsistency between platforms suggests no industry standard exists for handling such requests. OpenAI and xAI did not immediately respond to requests for comment about their policies on removing religious symbols from images.