What are the arguments and research evidence for and against the inclusion of AI in educational programmes?

Arguments against inclusion

  • Dependence on AI: There is apprehension that students may rely on AI as a replacement for their own thinking, potentially hindering their achievement of intended learning outcomes and the development of essential cognitive skills such as critical thinking. As discussed below, current research evidence does not support this concern. On the contrary, it points to a beneficial impact of AI use on cognitive skill development. Moreover, incorporating some activities on AI literacy in the classroom may help prevent students' overreliance on AI.
  • Erosion of fundamental knowledge: Critics argue that students might prioritize AI tools over acquiring foundational knowledge, which is crucial for their professional development and expertise. To prevent this, we advise fundamental knowledge and skills to be assessed in an AI-free environment AI and Assessment.
  • Assessment integrity: The use of AI in assessments raises significant concerns about the validity of evaluating whether students have achieved intended learning outcomes. This issue is particularly pressing for educators at UT, which is why we have dedicated an entire section of this resource hub to AI and Assessment.
  • Ethical considerations: The ethical implications of using AI in education are complex and warrant careful consideration, encompassing issues like bias, privacy, and accountability. We advise teachers to inform themselves about the ethical and privacy risks of using AI, see the hub section Fundamentals. We also advise curriculum developers wanting to incorporate AI literacy into the curriculum to include ethical considerations, see the section AI Literacy in Curricula.

Arguments for inclusion and supporting evidence

  • Personalised learning: AI can provide tailored educational experiences and personalised feedback, helping foster deeper understanding and engagement among students.
  • Development of higher-order skills: By using AI tools, students can engage in comparative analysis of their knowledge and reasoning, promoting reflection and critical thinking.
  • Development of domain knowledge: The use of AI tools as learning assistants (e.g., to automatically generate quizzes or practice questions) can help students acquire domain knowledge. At the same time, identifying misconceptions and gaps in AI-generated output may serve as a test for students’ knowledge.
  • Building AI literacy: Incorporating AI into curricula prepares students to use these tools responsibly, equipping them with essential skills for future workplaces where AI will likely play a crucial role. Considering Van den Akker’s (2013) famous metaphor, when the rationale at the centre of the curriculum ‘spiderweb’ changes, the curriculum should adapt accordingly.
  • Shaping responsible AI education: If we don’t teach students about AI, who will? It's crucial to approach AI education thoughtfully, rather than leaving the training of the future workforce to corporate interests.

Research Highlights

Research Study Information and Key Findings
Zheng et al. (2021) A meta-review of 24 articles published between 2001 and 2020 indicated a high effect size of AI on learning achievement.
Ouyang et al. (2022) A systematic review of 32 studies on the inclusion of AI in online education showed positive effects on learners’ engagement and academic performance.
Bond et al. (2024) A meta-review of systematic reviews (66 selected) on the impact of AI in higher education published between 2018 and 2023 identified the top 3 benefits as personalised learning, greater insight into students’ understanding, and positive influence on learning outcomes. In particular, AI chatbots (e.g., ChatGPT) were found to support student learning outcomes, critical thinking, empathy, communication skills, and satisfaction, as well as proficiency in a foreign language.
Essel et al. (2024) This study compared cognitive skill development in students using ChatGPT versus a control group, revealing a significant increase in critical thinking, reflective scepticism, and critical openness. The authors suggest that the personalized feedback from AI contributed to deeper subject understanding and skill development compared to a traditional lecture.