Given the rapid advances in AI and the momentum in the education field to understand how these technologies can support teaching and learning, last year the Gates Foundation launched a pilot initiative to provide funding to test new AI ideas that are in support of equitable K-12 mathematics outcomes. This is the second in a series of five blog posts elevating key learnings from this set of investments. This series will culminate in the release of a Digital Promise report on the cohort’s work in October. Today’s blog explores how three grantees embraced a culture of learning around AI. Check out last week’s post here.
Milton Rodriguez is blunt about the shortcomings of many AI-powered tools for use in the KIPP Chicago Public Schools, where he serves as senior vice president of innovation and development. “The design of some of the products and tools that we were reviewing did not include the insights of key members of our communities. Nor did they meet the needs of the communities that we serve,” Rodriguez reflects. “The technology is often designed for better resourced areas because those communities have a stronger voice in the shaping of those tools.”
In response to this reality, Rodriguez and his colleagues at KIPP Chicago Public Schools began to look for ways to center educators and students in the AI design and development process. For example, Rodriguez explains that KIPP Chicago teachers are serving more newcomer families and students than ever before. Yet available GenAI tools did not prioritize the needs of multi-language learners and their families. KIPP Chicago educators are currently working with industry partners to design GenAI-powered tools that offer educators better ways of communicating with families and students who may have been previously overlooked.
As part of our work with the Chicago Public Schools, KIPP Chicago Public Schools, and the Puerto Rico Department of Education, respectively, we led pilot projects to explore potential uses of AI by school districts as part of the foundation’s K-12 AI Pilot Cohort. We found that it was important to build a solid foundation for the use of GenAI by engaging staff, investing in learning with all stakeholders, and building bridges between departments before the rollout of an AI strategy. Learning and community buy-in are essential ingredients for success.
KIPP Chicago had not considered how AI tools could be used to better serve English learners and their families until a teacher raised it as a significant pain point. As we developed our K-12 AI pilot cohort projects, we all encountered situations where engaging staff led us to consider new approaches that could better serve school communities.
For example, through a research partnership with the University of Toronto, the Puerto Rico Department of Education’s project implemented the Spanish version of the AI-powered personalized tutoring tool Khanmigo in two schools. Through conversations with teachers, researchers learned that classrooms in the pilot lacked reliable internet, electrical power, and access to devices. Teachers and staff need to be able to use tools in their particular context to address their particular pain points. We need to talk directly to teachers and staff to understand those needs and challenges.
It is important to engage in learning with all members of your community and not rush to adopt policies or products. The GenAI landscape is constantly evolving, and the communities that we work with often have strong feelings about its use. For example, KIPP Chicago found that teachers had many concerns about AI’s impact on their careers, classrooms, and the field as a whole. Many community members were understandably concerned about student safety and safeguarding student-level data. KIPP Chicago held learning sessions where administrators, staff, and other stakeholders could together meet with industry partners and others to understand how the technology worked and to address those concerns. Community members’ needs, experiences, and concerns must be front-and-center as we develop district-level guidance around AI.
Further, school districts often serve diverse communities. Two schools in the same district may serve very different populations, have different pain points, and promote different approaches. We need to learn alongside members of all of those school communities and develop resources and guidance that are flexible enough to be useful in different contexts.
Chicago Public Schools developed an AI exploration framework to guide decision-making and an AI Guidebook based on its many internal learning and listening sessions with school communities throughout the district. The idea was not to prescribe a single approach or promote a particular set of tools, but rather to help school and district leaders and educators make decisions about how—and how not to—use AI. Chicago Public Schools has policies for data privacy and data integrations, as well as a list of approved edtech tools, many of which use AI. However, district leaders felt strongly that it was too early for a formal AI policy or the adoption or building of specific tools. First, the district needed to learn with those on the frontlines of schools and empower and support individual schools in serving their own communities.
Learning also needs to occur within district offices between departments that normally don’t collaborate, and between levels within a district. We need to create spaces where people from different positions and departments can come together to discuss values, concerns, strategies, and examples. The research partnership to study the implementation of the Spanish version of Khanmigo in Puerto Rico found that the centralized IT department lacked the resources to coordinate effectively with individual schools. The lack of coordination made it difficult to resolve technical challenges that teachers and students encountered in the Khanmigo implementation.
Chicago Public Schools’ AI steering committee shows how coordination between departments can work in a large district. The AI steering committee is led by the Office of Teaching and Learning and Information and Technology Services—two offices representing many different departments. The district is currently applying the AI exploration framework to determine if and how to use AI to enhance Skyline, its existing digital curriculum. Because the team that created the framework was cross-functional, district leaders believe that the decisions will better reflect the views of all stakeholders, taking into account legal, technical, and instructional viewpoints.
These are still early days for GenAI, and it will take time for districts to develop policies and strategies that reflect the varied needs and priorities of their communities. Engaging staff, investing in learning, and building bridges between departments can help build a solid foundation for meaningful AI guidance and decision-making.
Be on the lookout for Digital Promise’s full report on the K-12 AI Pilot Cohort, to be published in October.
Want to know more about how school districts are embracing a culture of learning in developing AI strategies? Find more resources here:
