AI Doing Homework? Teachers Are Bringing Students Back to In-Class Exams

A new study by HIT and the Hebrew University, published in the journal Teaching and Teacher Education, reveals that many middle school teachers in Israel are restricting the use of artificial intelligence (GenAI) in homework assignments, preferring in-class exams, classwork, and oral presentations to ensure that assessment reflects students' true knowledge. The researchers call this approach 'pedagogy of blocking'. The study is based on interviews with 24 teachers and points to uncertainty and heavy responsibility placed on teachers in dealing with the change.

A new study by HIT – Holon Institute of Technology and the Hebrew University of Jerusalem, published in the journal Teaching and Teacher Education, reveals how artificial intelligence is changing assessment methods in schools. The researchers – Dr. Meital Amzaleg and Prof. Gila Kurtz from HIT, and Omri Hader and Prof. Yifat Ben-David Kolikant from the Hebrew University – interviewed 24 middle school teachers in Israel. Nearly 90% of the interviewees described situations where they restrict the use of GenAI, even teachers who integrate the technology into their teaching. They prefer classwork, exams instead of home projects, and oral presentations, to ensure that assessment reflects the student's own knowledge. The researchers call this consideration 'pedagogy of blocking': the teacher examines whether the use of AI advances the learning goal or allows bypassing the action the student is supposed to learn. Examples: a literature teacher considers replacing home projects with in-class exams; an English teacher uses ChatGPT for speaking practice but restricts use in independent writing. The study identifies four thinking patterns: developing higher-order thinking, maintaining authority, imparting knowledge and information credibility, and flexibility. 15 teachers raised dilemmas, 17 felt heavy responsibility, and 11 described a sense of loneliness. The researchers emphasize that this is a qualitative study intended to identify patterns, not to determine prevalence.