AI & RoboticsMore in AI & Robotics→
Two-Year Study Finds Banning AI From Classrooms Leaves Students Worse Off
Key Takeaways
- Law professor Thibault Schrepel of Vrije Universiteit Amsterdam ran the same three-group experiment in 2024 (66 students) and 2025 (164 participants) in his Law of AI course.
- Students barred from ChatGPT finished last both years, making mostly superficial wording edits and hitting what the researcher calls 'idea exhaustion' after 10-15 minutes.
- The unguided AI group accepted ChatGPT revisions largely without question and every subgroup kept at least one misleading or legally extraneous term.
- The trained group, taught legal prompt engineering and accuracy checking, scored far above the others in 2024, but the advantage nearly vanished in 2025 as chatbot familiarity grew.
DE
DT Editorial Team··via the-decoder.com