AI Homework Use Boosts Scores 18%, Then Drops Exam Results 20%

A study involving 26,811 Chinese students in grades seven through 12 has revealed a significant dichotomy in the impact of AI on academic performance. Researchers from the Centre for Economic Policy Research found that AI adoption led to an 18% increase in homework scores and a 30% reduction in completion time. However, this initial boost was short-lived, as monthly exam scores declined by 20% within six months of AI integration. The negative trend continued, with college entrance exam scores falling by 18% to 24%, reaching their lowest point after two years.

The research, conducted by academics from Stockholm University and the University of Hong Kong, identified a specific student profile most affected by these diverging outcomes: those who "outsourced" their homework by using AI to complete assignments accurately and quickly. Approximately 80% of these students experienced the detrimental effect on their test scores, both in the short and long term. This finding highlights a disconnect between perceived AI productivity and actual learning gains, fueling concerns among educators and experts regarding the widespread, unmonitored use of AI in education.

Researchers emphasized that the primary objective for students should be learning from tasks, not merely completing them efficiently. The rapid adoption of generative AI tools among students has consequently raised significant concerns about their learning processes. The study's findings suggest that generative AI, poised to become a pervasive educational technology, carries a substantial negative impact on student learning outcomes. This research adds empirical evidence to ongoing discussions about the critical engagement of younger generations with educational materials, particularly as AI adoption becomes increasingly synonymous with academic shortcuts.