The emergence of AI brings new solutions to old problems. It also creates some unique problems of its own. Beyond concerns over privacy, bias, and reliability, AI is unleashing a deluge of new products across the board-education, no exception. As options multiply, districts and schools are asking how to sift through the effective from the cleverly marketed and boldly promised.
As part of the federally funded LEARN Network that supports the development and scaling of quality educational products and programs, I have worked alongside hundreds of researchers, developers, practitioners, and educators across the country to help bring their best ideas to the field. Together, we have gained unique insights into what makes some products fail while others succeed, and we have learned valuable information that districts and schools might consider when making decisions about new tools and programs.
Effective edtech should never be created to replace human relationships with students.
One thing we have learned is that this wave of AI-based edtech is not so different from what we are used to. Some products show a lot of promise, and some fall flat. The choices are more plentiful the technology sophisticated, and watchfulness in selection processes must remain a priority for schools. Based on our work and our conversations with leaders in this space, here are some important questions to ask while searching for an edtech solution in the age of AI.
What does it do?
Effective edtech was never and should never be designed to substitute for human relationships with students. State policymakers in California and Minnesota and organizations like the National Education Association are responding to new school-based AI programs by pushing back to make sure that educators are always front and center in education. Quality edtech, whether powered by AI or not, should enhance education outcomes and efficiency.
While certainly one key difference concerns student-facing AI versus products and programs intended for practitioners, administrators, and other staff, both uses raise unique considerations. In student-facing products, guardrails on development need to ensure that AI is developed to be free of bias, privacy protections are in place, and reliance on the product is highly reliable. For administrative applications, considerations will probably center on whether the edtech increases efficiency yet leverages human expertise.