Teachers can edit an AI draft-or decide it does not fit
A revised preprint reports interviews with 30 frequent AI-using teachers in South Africa, Taiwan and the US. Some AI drafts needed small edits; others had gaps that prompting could not fix. The interviews describe teacher choices, not a measured district outcome.
What the source reports
Qing Xiao and colleagues interviewed 30 K-12 teachers who often use AI. The teachers worked in South Africa, Taiwan and the United States. The paper describes different levels of work needed to fit AI output to a real lesson: a small edit, a major rewrite or a case where the draft cannot be used. That is a practical reason to check local language, materials and learner goals before sharing a draft. This is a revision of an earlier exploratory preprint, not a new field trial in 2026. It does not prove a district-wide benefit or tell any school to put student records into an unapproved tool.
Original source
- Title
- How K-12 Teachers Adapt Generative AI to Local Classroom Demands: Cultural Distance Between GenAI Defaults and Everyday Teaching Practices
- Author
- Qing Xiao, Ruiwei Xiao, Hanqi Jane Li, Phenyo Phemelo Moletsane, Xinying Hou, John Stamper, Hong Shen
- Publication
- arXiv
- Date
- Tuesday, October 27, 2026