Researchers found that students who used generative AI for homework and exam preparation scored significantly lower on independent problem-solving assessments compared to peers who relied on traditional study methods.
How AI use affects cognitive development in educational settings
The study tracked over 1,200 secondary school students across Germany over six months, measuring changes in analytical reasoning and memory retention through standardized tests before and after introducing AI-assisted learning tools.
Participants who used AI for more than 30% of their study time showed a 22% decline in unaided reasoning scores, while those using AI less than 10% of the time maintained or improved their baseline performance.
Why educators are reconsidering AI integration in classrooms
Education policymakers in several German states have paused planned expansions of AI tutoring programs pending further review, citing concerns about long-term skill atrophy rather than short-term efficiency gains.
Previous pilot programs in Bavaria and North Rhine-Westphalia had reported improved task completion speeds but did not assess deeper cognitive outcomes, creating a gap now highlighted by this research.
What this means for future learning environments
Experts suggest that AI tools may need redesign to prompt rather than replace thinking, such as requiring students to justify AI-generated answers before submission, to preserve cognitive engagement.
Does this mean AI should be banned in schools?
No, the researchers emphasize that AI can support learning when used as a supplementary tool, but warn against overreliance that undermines the development of independent thought.
Are younger students more affected than older ones?
The study did not break down results by age group, so no conclusion can be drawn about differential impacts across school levels based on the available data.
