Generative artificial intelligence (AI) platforms have upended common methods for evaluating student learning. Traditionally, college courses require students to work on their own and then submit final products of learning for evaluation. However, because of the capabilities of AI, teachers can no longer trust that student work reflects the achievement of traditional learning goals. When trust in psychology’s empirical results waned, the solution was to make the process of achieving those results transparent through open science practices. Analogously, when teachers cannot trust academic work to represent student learning, the solution is to adopt open learning practices that require students to document the work they put into the learning process. Open science includes preregistration of study plans. Similarly, teachers can establish a shared plan for learning with students before they start a project. Open science includes data and analysis transparency. In open learning practices, there is a definition of acceptable sources, including AI. Also, students share copies of their sources and document use of sources through methods such as shared annotations. The open science practice of sharing study protocols clarifies research procedures, and the open learning practices equivalent is for students to show work through shared drafts and revisions. Finally, open science follows set standards in the reporting of study results, and the open learning practices equivalent is to cite all sources, include AI contribution statements, and share work portfolios. By opening up student work to direct observation, teachers can better ensure that their evaluations of learning have integrity.
