AI headshot apps removed her hijab. A Berkeley Law researcher wants to know why.

Mahwish Moazzam, a J.S.D. candidate at Berkeley Law, discovered that over 25 AI headshot generators consistently removed her hijab, raising concerns about AI bias and systematic alteration of visible religious identity.
Moazzam conducted her tests over a year, finding that none of the apps asked if the hijab should be retained, while some offered to keep other accessories like glasses.
This issue of visible identity modification underscores broader questions about discrimination law, representation, and human dignity as algorithms may erase or alter visible indicators of identity like religious attire.
The findings suggest potential legal and human rights challenges as accountability can become complex for an AI generated harm, especially when crossing international borders.
Moazzam’s research highlights a less explored form of algorithmic bias, emphasizing that as AI technologies proliferate, they may reproduce hidden biases that current laws do not address effectively.
The professor and peers at Berkeley Law appreciate the societal and legal implications of her work and emphasize the need for new legal frameworks to address AI-generated harm and the challenges it poses to human rights and dignities.