A qualitative interview study exploring when generative AI supports and fails everyday K–12 teaching globally

Generative AI (GenAI) is increasingly being used in K–12 education, yet its value in everyday teaching remains uneven across tasks and contexts. While prior discussions often focus on potential benefits or risks, less is known about when GenAI meaningfully supports teachers’ daily work and when it fails to do so. This study aimed to examine how K–12 teachers use GenAI in routine teaching practices and to identify areas where current tools align with, or fall short of, classroom needs. Qualitative data were collected through semi-structured interviews with 30 teachers from South Africa, Taiwan, and the United States. Teachers described a range of uses, including communication with parents, classroom activity design, assessment preparation, and culturally specific instruction. We found that GenAI supported some tasks with minimal effort, while other tasks required substantial teacher intervention, and some could not be effectively supported due to language limitations or policy restrictions. These findings suggest that GenAI currently offers selective benefits while shifting varying amounts of additional work onto teachers. Designing tools that better align with everyday teaching practices and local educational contexts may improve adoption and reduce hidden labor in classrooms.