Frequent use of AI may hinder students' academic performance (phys.org)

A study by researchers at the University of Tartu investigated how frequent use of AI tools, such as ChatGPT, affects students’ academic performance in an object-oriented programming course. Surveying 231 students, the researchers found that while AI was primarily used for debugging and understanding code, increased reliance on these tools correlated with lower academic results. This suggests that unguided and excessive use of AI may hinder learning rather than support it, possibly because struggling students turn to AI as a crutch rather than a supplement. The findings highlight the importance of integrating AI thoughtfully in education—emphasizing its role as a supportive aid rather than a shortcut. Associate professor Marina Lepp stresses that proper guidance is crucial for fostering critical skills and ensuring AI enhances rather than replaces the learning process. The study also noted positive uses of AI, such as creatively translating code between programming languages, which can facilitate learning new skills. These insights are valuable for educators seeking to balance AI’s benefits with its potential drawbacks, shaping effective strategies to incorporate AI in teaching while maintaining educational rigor.