Screen-Shot-2024-07-01-at-9.44.37-AM

Diversity and Inclusion in Artificial Intelligence Development

Authors: Erika Moore, Ph.D.; info@ciddl.org

Ensuring diverse representation in discussions and developing artificial intelligence (AI) in education is crucial. This diversity is essential for AI to effectively bridge educational gaps for students with disabilities and other academically struggling populations. Including thought leaders and experts from diverse backgrounds is key to integrating AI into the classroom effectively and safely.

A Changing US Student Population

Demographic changes present the need for inclusive AI in education. By 2050, non-Hispanic White children are projected to comprise just 42% of the school-aged population in the US, with Hispanic, Black, Asian, Pacific Islander, and multiracial children making up the majority (Van, 2023). Additionally, about 15% of school-age children have an identified disability and are entitled to an Individual Education Program (IEP) under the Individuals with Disabilities Education Act (NCES, 2024). This shift necessitates AI education tools that serve a more diverse student body. Therefore, AI systems must be designed to account for various cultural backgrounds and languages, providing personalized learning experiences that are culturally sensitive and linguistically inclusive.

Building Diverse Teams to Address Bias

Multidisciplinary teams, including technologists, ethicists, and other professionals, are essential for AI development. However, ensuring these teams are diverse is even more critical when developing AI tools for education. According to an executive summary on Diversity in High Tech published by the U.S Equal Employment Opportunity Commission (EEOC), around 63% of employees in the U.S. technology sector are White, 20% identify as Asian American, 8% as Hispanic or Latino, 7% as Black, and less than 1% as Native Hawaiian, Pacific Islander, or Native American. These statistics highlight the importance of amplifying minority voices in the technology sector.

Diverse teams bring various perspectives and experiences that are vital in identifying and mitigating biases ensuring AI systems are fair, inclusive, and effective. Data quality and diversity are crucial for AI performance and fairness when using generative AI systems such as ChatGPT, Copilot, and Gemini. Datasets must be designed ethically and with equity in mind. Diverse teams are better equipped to recognize and address biases in AI systems (Shams, 2023). A team of diverse professionals can identify sources of bias that homogeneous teams might overlook, ensuring AI systems do not perpetuate existing inequalities. Assembling diverse teams to develop AI tools for education ensures all voices are heard and valued, which is core to considering students of all races, ethnicities, genders, and abilities.

Ethical AI for Education

Educational institutions, research organizations, AI developers, and other stakeholders must prioritize diversity, equity, and inclusion throughout the construction of educational technologies for students. This includes establishing diverse expert teams to create ethical guidelines and ensuring continuous oversight and adaptation in AI practices. Incorporating continuous feedback loops from diverse stakeholders to iteratively improve the fairness and accuracy of AI systems in education is essential (Shams, 2023). Creating inclusive AI tools, addressing biases, ensuring equitable access, and preparing all students for success in an increasingly diverse and technologically advanced society requires collective intentionality. 

Keep the Conversation Going

How well do you think AI educational technology addresses diversity and inclusion in their programming? Please share your thoughts in our community.

References 

National Center for Education Statistics. (2024). Students With Disabilities. Condition of Education. U.S. Department of Education, Institute of Education Sciences. https://nces.ed.gov/programs/coe/indicator/cgg.  

Shams, R. A., Zowghi, D., & Bano, M. (2023). AI and the quest for diversity and inclusion: a systematic literature review. AI and Ethics, 1-28.

U.S. Equal Employment Opportunity Commission (EEOC). (2014). Diversity in high tech. https://www.eeoc.gov/special-report/diversity-high-tech

Van Hook, J., Bélanger, A., Sabourin, P., Patoine Hamel, N. (2023). The changing racial and ethnic composition of the school-age population in the U.S. Los Angeles, CA: The Civil Rights Project/Proyecto Derechos Civiles, UCLA.