If you can't beat them, join them: Incorporating AI into Higher Education assessments to explore the perpetuation of bias, underrepresentation, and stereotyping.

Conference: Society of Legal Scholars (SLS) international Symposium From Binary to Bench – Empowering Equal Opportunities and AI Literacy in Global Legal Education

Dates: 5 November 2025

Abstract:

As a feminist educator highlighting and dismantling in built societal bias and stereotypes in media is a key part of my research and pedagogy. The proliferation of Generative AI usage amongst students has caused more than worries about a lack of engagement, but also the very real danger of perpetuating and reinforcing existing harmful norms. Generative AI has been found to reflect and amplify biases and thus leads to “exacerbating inequalities and discrimination” Ulnicane and Aden (2023 p. 2). It would be naïve to suggest that students are not engaged in using Generative AI within even the most innovative assignment designs. With this in mind this paper explores the implementation of using Generative AI to analyse the inherent biases built into such systems, to inform students of the theoretical underpinnings of media bias, representation theory, and hegemony and to ask them to use Generative AI to explore how they may be perpetuated. The assignment then encourages an in person, creative element to enable students to design and produce their own representative images to counteract intrenched media bias and to identify hallucinations formed from bias datasets (Burger et al 2023). The aim of this assessment type is to encourage critical use of Generative AI and to encourage critical analysis as a media consumer. By highlighting the limitations of Generative AI so too is it hoped that this will deter students from over reliance on it as a writing tool.

Source: Manual