Impact of AI on (higher) education
Introduction
The scientific literature on the impact of AI on learning processes is growing rapidly. We distinguish three domains: cognitive impact, motivational effects and social dimensions of learning.
Cognitive impact
Cognitive offloading
A recurring theme in the literature (Risko & Gilbert, 2016; Sweller, 2023) is cognitive offloading: the delegation of mental tasks to external resources. AI significantly amplifies this phenomenon.
Risks:
- Reduced development of basic skills (writing, arithmetic, reasoning)
- "Automation bias": blind acceptance of AI outputs
- Reduced memory usage and deep processing
Opportunities:
- More cognitive space for higher-order thinking
- Faster iteration and experimentation
- Scaffolding for weaker students
Zone of Proximal Development
Vygotsky's concept of the Zone of Proximal Development (ZPD) has been reinterpreted in AI context in recent literature. Chan & Hu (2023) argue that AI can provide a unique form of scaffolding that adapts to the student's level - something that is difficult for the teacher to achieve in large groups.
Motivational effects
A meta-analysis by Zawacki-Richter et al (2022) across 146 studies shows a mixed picture:
- Students report higher self-efficacy when AI provides direct feedback
- At the same time, intrinsic motivation decreases when AI takes over tasks completely
- Achievement-oriented students show stronger tendency to over-reliance on AI
Social dimension
Learning is inherently social (Wenger, 1998). The literature warns of the risk of AI use undermining the social dimension of learning:
- Less peer interaction in AI-assisted assignments
- Risk of "AI isolation": student solves problems with AI rather than consulting fellow students
- Language acquisition research shows reduced output among students using AI for language production (Huang et al., 2024)
Academic integrity
This is the most discussed topic in recent education literature. Thurlings & Zijlstra (2023) distinguish three positions:
| Position | Starting point | Implication for policy | . |
|---|---|---|---|
| Ban approach | AI = fraud | Detection tools, explicit bans | |
| Transition approach | AI = new reality | Adjust assignments, allow AI if justified | |
| Integrative approach | AI = opportunity | explicitly incorporate AI use as learning activity |
The international consensus is shifting towards the integrative approach, where AI use is used transparently and purposefully as part of the learning process.
References
- Chan, C. K. Y., & Hu, W. (2023). Students' voices on generative AI. Computers and Education: Artificial Intelligence, 5.
- Huang, J., et al. (2024). The impact of AI writing assistants on L2 writing development. Language Learning & Technology, 28(1).
- Thurlings, M., & Zijlstra, H. (2023). Academic integrity in the age of generative AI. Higher Education, 86, 1-18.
- Zawacki-Richter, O., et al. (2022). Systematic review of AI in higher education. International Journal of Educational Technology in Higher Education, 19(1).