Does AI actually improve learning, or just make students feel busier?
The strongest evidence here comes from a 2025 study at Ghana Communication Technology University, which surveyed 185 students and found a positive correlation between AI tool use and student engagement — meaning students who used AI-driven methods were more actively involved in their studies [1]. However, the same study found that AI's effect on actual academic performance was weak, suggesting that engagement alone doesn't translate into higher grades [1]. This is a critical distinction: AI policies justified by claims of better learning outcomes may overpromise, but they can be justified by the real benefit of keeping students engaged.
Do students already know how to use AI ethically?
A 2023 survey of 201 Croatian university students found that more than half use ChatGPT for written assignments, most commonly to generate ideas, but also to summarize, paraphrase, proofread, and even write parts of assignments for them [3]. Crucially, students themselves rated idea generation as the most ethically acceptable use, while other uses were seen as unethical — yet some students still engaged in behaviors they knew were wrong [3]. This gap between knowing and doing is exactly the kind of problem a clear policy can address, by setting firm boundaries and consequences.
What actually makes students adopt AI — and how can policies shape that?
A 2026 study of 441 business students across the Baltic Sea region found that the strongest drivers of generative AI adoption were habit, social influence, and perceived study value — not performance expectations or ease of use [2]. This means students adopt AI because it becomes routine and because their peers do it, not because they carefully weigh its academic benefits. For policymakers, this suggests that simply telling students AI is useful won't change behavior; instead, policies should focus on creating positive peer norms and habitual, ethical use from the start [2]. The same study found no significant effect of 'facilitating conditions' (like institutional support), which implies that policies alone won't drive adoption unless they also shape the social environment.
About These Sources
This answer is built on 3 peer-reviewed studies — published from 2023 to 2026, 2 from 2024 or later, 1 in Q1–Q2 journals, collectively cited 63 times — selected as the most relevant from 3 studies that passed quality screening, drawn from 54 papers retrieved from a database of over 500 million.
Sources used in this answer
The Role of AI Technology in Enhancing Learning Outcomes of University Students: A Case Study of Ghana Communication Technology University (GCTU)
In a survey of 185 university students in Ghana, AI integration showed a positive correlation with student engagement but only a weak effect on academic performance, suggesting AI boosts involvement more than grades.
Generative AI Adoption Among Business School Students: Drivers of Use and Policy Implications
A survey of 441 business students across the Baltic Sea region found that habit, social influence, and study value were the main drivers of generative AI adoption, while performance expectancy and ease of use were not significant.
Writing with AI: University Students’ Use of ChatGPT
A survey of 201 Croatian university students found that over half use ChatGPT for assignments, mostly for idea generation, but many engage in uses they consider unethical, revealing a gap between ethical judgment and behavior.