The gendered dress of DALL-E 2: Exploring profession-based images in the Indian context

Abstract

This study delves into the intricate realm of gender and artificial intelligence (AI) through an examination of DALL-E 2 generated images within the Indian context. The study takes a methodological approach that focuses on assessing images generated in response to prompts such as ‘a farmer cultivating crops in rural Punjab’ or ‘a nurse providing care in a hospital in Delhi’ to reveal the dynamics of gender performativity within profession-based visual content. The generated images were analysed using the dress as a phenomenon to visualise Indian man, Indian woman and ambiguous Indian. The study concludes that DALL-E 2’s algorithm reiterates the binary gender norms leaving no space for ambiguous Indian in its responses. Though the generated image is centred on female professions, it is largely employed by male surroundings. Hindu religious symbols are largely used among male professionals denoting its predominance. In its aesthetics, the individuals portrayed exhibit a brown complexion, echoing the demographic landscape irrespective of gender.