Written by Leah Dozier Walker
Executive Vice President of Equity & Inclusion at Waterford.org
The integration of Artificial Intelligence (AI) holds tremendous promise across the education ecosystem. It has the potential to revolutionize learning experiences, enhance family engagement, and drive academic achievement. Leveraging AI can power differentiated instruction, personalize learning pathways, develop racially and culturally inclusive content, and provide invaluable feedback to educators and administrators. But like other innovations in education, it cannot be excellent if it is not inclusive.
AI education tools must be designed with the goal of preparing learners for success in an increasingly diverse and global society. According to a recent 2023 UCLA study, by 2050, non-Hispanic White children are projected to compose just 42% of the school-aged population (ages 5-17), Hispanics will represent 29%, Blacks will represent 17%, Asians and Pacific Islanders will represent 7%, and children with multiracial or other identities will represent 4%.
These demographic shifts necessitate a conscious and deliberate effort from both solution providers and users to build inclusive and diverse teams to scrutinize the quality of data being utilized, including the acquisition of datasets, the application of their products, and the monitoring to avoid any inadvertent biases.
Nothing about us, without us, is for us
AI education tools will be leveraged to prepare students for success in an increasingly diverse and interconnected world. As we delve into the realm of AI in education, it’s crucial to approach this new frontier with an inclusive lens and ethical scrutiny. While the benefits are enticing, we have an ethical responsibility to ensure that we do not perpetuate existing biases and inequities in our education systems.
One of the most significant pitfalls would be to rush forward without considering diverse perspectives and involving non-traditional stakeholders as co-creators of these innovations. This can be mitigated by assembling diverse teams of developers, data scientists, and subject matter experts from various cultural and ethnic backgrounds. Utilizing diverse perspectives can help identify and address biases in the dataset collection process.
Mitigate bias through accountability
AI technologies, like any tool, are not neutral. They reflect the biases, both intentional and unintentional, of their creators. There’s a legitimate concern that without careful oversight, AI will only become “better” at reproducing bias, systemic racism, and discrimination.
To address this, decision-makers must demand evidence of culturally diverse and ethnically accurate datasets. These datasets are the foundation upon which AI resources are built, and they must be designed ethically and with equity in mind. Additionally, developers must incorporate continuous feedback loops from a diverse pool of stakeholders to iteratively improve the fairness and accuracy of AI systems.
The way we develop and implement AI in education holds immense significance for individuals in the most marginalized communities. The biases inherent in algorithmic tools and their underlying datasets stem from their human creators. Humans, influenced by societal norms, are conditioned to make judgments based on identity markers such as race, gender, class, immigration status, and religion. Failure to intentionally assemble racially and culturally diverse AI product development teams poses a considerable risk that these tools will perpetuate and even exacerbate societal inequalities.
At Waterford.org we will forge this new frontier with the same commitment to inclusive excellence that is foundational to our content development, product design, and program delivery. We are leveraging collaboration with national experts, including our National Advisory Council on Inclusive Practices, to ensure that diverse perspectives and communities inform our AI innovations. Finally, we are intentionally seeking out voices and perspectives from outside tech spaces to ensure that we are aware of bias in our data and the outcomes it generates.
In the pursuit of harnessing the potential of AI in education, we must not overlook the necessity of embracing diverse perspectives, demanding accountability, and prioritizing equitable and inclusive practices. We can and must ensure that AI serves as a force for positive change in our educational landscape. Together, let’s strive to build a future where AI enhances learning opportunities for all, without perpetuating the biases and systemic inequities impacting marginalized communities today.
Leah Dozier Walker serves as the Executive Vice President for Equity and Inclusion at Waterford.org, a national early education not-for-profit whose programs reach over 300,000 children in 43 states each year. At Waterford, she leads the development and implementation of inclusive excellence strategies both internally and externally, while aligning Waterford’s organizational priorities to promote inclusive excellence across all operational domains.
She is the founder and serves as principal consultant at Modern Impact Solutions, a professional services practice providing an array of inclusive excellence, communications, and strategic advising consultation to corporate, non-profit, education, and public sector organizations and executives. Prior to this, she served as the first-ever Equity Director at the Virginia Department of Education where she spearheaded education policy development and professional learning programs to advance culturally responsive educator practices and disrupt disproportionate student outcomes.
