Article citationsMore>>

Krishnan, A., & Rattani, A. (2023). A Novel Approach for Bias Mitigation of Gender Classification Algorithms Using Consistency Regularization. Image and Vision Computing, 137, Article ID: 104793.
https://doi.org/10.1016/j.imavis.2023.104793

has been cited by the following article:

• TITLE: Fairness in AI: When are AI Tools Gender-Biased?

  AUTHORS: Dialekti Athina Voutyrakou, Gianna Katsiampoura, Constantine Skordoulis

  KEYWORDS: Artificial Intelligence, Gender Bias, ChatGPT, Gemini, Ethical AI

  JOURNAL NAME: Advances in Applied Sociology, Vol.15 No.3, March 12, 2025

  ABSTRACT: In the last decade, the implementation of different Artificial Intelligence (AI) tools has rapidly increased and found application in several domains, e.g., healthcare, education, autonomous driving etc. Therefore, ensuring fairness in these systems has received increasing attention, becoming a topic of considerable interest over the last few years. This work focuses on gender bias within AI tools, by providing a comprehensive literature review that examines the causes and implications of this bias and presenting the findings of four experiments that the authors conducted within two widely used AI tools, namely ChatGPT 3.5 and 4.0 and Gemini 1.5 and 2.0 Flash. These experiments aim to investigate the potential presence of gender-biased responses, as well as the influence of sociocultural norms on the outcomes of these AI tools. Following this, the results are analyzed and mitigation strategies along with policy recommendations are proposed to support the development of Gender Unbiased AI tools.