Learning Commons and Cricket Media are collaborating to help organizations evaluate the complexity and grade-level appropriateness of AI-generated texts for classrooms. AI presents opportunities to improve the efficiency of educational content development, and our organizations are focused on building tools to ensure it’s done with appropriate guardrails. Founded in 1973 and recognized with prestigious honors, including the Parents’ Choice Awards for excellence in children’s publishing, Cricket Media produces high-quality literary and informational curriculum content and supports educational publishers with its AI-enabled CricketSafe platform. Through this collaboration, Cricket Media will contribute approximately 500 reading passages from its Spider and Cricket magazine libraries to support Learning Commons’ open-source Evaluators. Learning Commons is collaborating with Student Achievement Partners to annotate the passages for its qualitative text complexity Evaluator suite. By reviewing and expert-labeling these passages, Student Achievement Partners provides the critical data Evaluators need to ensure AI-written classroom texts meet rigorous, research-based standards for grade-level appropriateness.
As AI education tools proliferate, educators and developers require clear, credible evaluation tools to assess whether these technologies support meaningful learning. By anchoring Evaluators in established educational research and grade-level-appropriate reading content, Learning Commons makes it possible to better understand text complexity and assess how AI-generated passages compare with research-grounded criteria.
Improving evaluation based on high-quality content
Evaluating AI for education requires more than measuring technical performance. It also means assessing whether content reflects sound pedagogical practices and supports teaching and learning. Open, transparent evaluation tools can help developers make more informed decisions while giving researchers, educators, and other stakeholders the opportunity to examine, adapt, and build on shared approaches.
The reading passages from Cricket Media include engaging literary and informational texts spanning topics in science, history, nature, technology, and the arts. The passages offer opportunities to examine text complexity, including language and structure, as well as the background knowledge and vocabulary readers need to comprehend a text. This partnership reflects how organizations across the education ecosystem can contribute high-quality resources to shared, open-source evaluation tools that help advance the quality of AI-generated educational content.
Looking forward
Learning Commons will continue working with partners across education to translate research and proven teaching and learning practices into openly available tools, data, and frameworks. By making rigorous text complexity evaluation resources accessible to the broader ecosystem, Learning Commons aims to help advance AI-powered tools that better serve educators and students.