Gender biases in AI - Mitigation strategies contributing to fairness
Ahonen, Eija; Farén, Eeva (2024)
2024
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2024052314820
https://urn.fi/URN:NBN:fi:amk-2024052314820
Tiivistelmä
Humanity is entering the age of digitalisation, marked by rapid advances in technologies such as AI. In the midst of this transformation, ensuring gender equality and fairness in the digital sphere is paramount. Traditional views of the benefits of technology have focused on cost reduction and productivity gains, but the rise of big data has highlighted the importance of intelligent technologies for efficient data analysis. However, unchecked assumptions about the rationality of AI have led to the misconception that it can reliably handle morally sensitive decisions, raising ethical concerns about bias and fairness.
Gender bias studies reveal discrimination against women in various AI applications, from translation tools to recruitment processes. The underrepresentation of women in ICT jobs exacerbates the challenges and risks for women in the digital age. Digital gender equality initiatives have gained traction globally, highlighting the need to address gender bias in AI.
This thesis is conducted as a systematic literature review to collect and evaluate research on gender bias in AI, focusing on mitigation strategies to address risks for women. Through qualitative analysis, it aims to elucidate the underlying mechanisms of gender bias in digital technologies and to assess the effectiveness of mitigation strategies in promoting equity and fairness. By contributing to ongoing discussions in the field, this research aims to promote a more equitable and inclusive technological framework, driven by the vision of a technologically advanced society based on gender equality.
Gender bias studies reveal discrimination against women in various AI applications, from translation tools to recruitment processes. The underrepresentation of women in ICT jobs exacerbates the challenges and risks for women in the digital age. Digital gender equality initiatives have gained traction globally, highlighting the need to address gender bias in AI.
This thesis is conducted as a systematic literature review to collect and evaluate research on gender bias in AI, focusing on mitigation strategies to address risks for women. Through qualitative analysis, it aims to elucidate the underlying mechanisms of gender bias in digital technologies and to assess the effectiveness of mitigation strategies in promoting equity and fairness. By contributing to ongoing discussions in the field, this research aims to promote a more equitable and inclusive technological framework, driven by the vision of a technologically advanced society based on gender equality.