The Potential of Diverse Youth as Stakeholders in Identifying and Mitigating Algorithmic Bias for a Future of Fairer AI

Introduction

With trends revealing the general populace’s growing use of Artificial Intelligence (AI), it is becoming more evident that we, as a society, are becoming dependent on AI.   Given the increased use of AI, it is important to note that AI technologies also possess inherent social and cultural biases that reflect and perpetuate existing societal problems. Young people, ages 12-18, are some of the most frequent users of AI; however, they are not engaged in discussions centered on AI’s biases, the real-world implications of those biases, and the ethics of AI systems. 

This paper examines how young people engage with artificial intelligence, specifically the interpretation and management of algorithmic biases. The research challenges a common narrative that young people are only consumers of technology who do not have the knowledge and interest to relate to technology’s moral or technical implications. This research proposes that young people are capable of thinking about the consequences of using AI from different perspectives. Researchers were interested in knowing what young people think about AI, identifying AI-associated biases, and suggesting ways to reduce these biases,  which would foster fairer AI technologies.

Jaemarie Solyst is a PhD candidate at Carnegie Mellon University affiliated with the Human-Computer Interaction Institute. Ellia Yang is an undergraduate student studying computer science and human-computer interaction at Carnegie Mellon University.  Shixian Xie is a PhD student at Carnegie Mellon University affiliated with the Human-Computer Interaction Institute. Amy Ogan is an associate professor of learning sciences affiliated with the Human-Computer Interaction Institute at Carnegie Mellon University. Jessica Hammer is an associate professor and the director of the Center for Transformational Play at the Human-Computer Interaction Institute at Carnegie Mellon University. Motahhare Eslami is an assistant professor affiliated with the Human-Computer Interaction Institute at Carnegie Mellon University.

Methods and Findings

The participants included students(from first through twelfth grade) and their parents,  primarily from backgrounds that are underrepresented in technology, including Black communities, non-binary users, and women. The participants engaged in a series of interactive workshops and discussions exploring the concepts of bias and discrimination in technology, examined case studies, and critically analyzed social media platforms. The study found that youth were able to explain their understanding of bias and discrimination on social media platforms, including how algorithms perpetuate these biases.

Furthermore, participants discussed strategies for mitigating and eliminating bias, as well as advocating for more equitable AI. Drawing on their experiences and insights, the participants framed the concept of fairness within the context of algorithms. Their solutions, which blended creativity with practicality, showed that they understood the ethical dilemmas in AI. The participants proposed various recommendations, from technical fixes like strengthening bias detection algorithms to broader societal and educational initiatives such as developing comprehensive AI education programs and promoting public dialogue on algorithmic accountability.

Ultimately, these findings contradict the belief that young people possess no interest or ability to confront complex moral and technological dilemmas. Instead, this study illuminates an opportunity for ways young people can become active participants in building or critiquing AI technologies. The study participants’ insights could lead to more equitable and thoughtful development processes for AIs.

Conclusions

The study underscores how young people can help shape AI, and argues that with the inputs of youth, fairness and inclusivity of AI will be protected, advocating for increased participation of young people in AI design. By actively involving youth in the discourse around AI ethics and development, their unique perspectives and innovative ideas challenge biases and ensure that AI systems are more representative of and responsive to the diverse needs of society. This study highlights the importance of educational programs that help empower young people with tools for critically examining, evaluating, and contributing to AIs, thereby fostering an informed and ethically aware generation engaged in creating or interacting with such technology.