The rapid rise of Generative AI (GenAI) tools, such as ChatGPT, Claude and Gemini, has transformed many aspects of our world, including education. These tools have sparked significant debate, particularly concerning academic integrity. For teachers, the sudden (and often unguided) integration of AI into learning raises important questions: How can we ensure fair assessments when students have access to advanced AI tools? Is it possible to turn these technologies into allies rather than adversaries in the classroom?
Challenges of GenAI in Assessments
Cheating is not a new issue in education; it has existed for decades, driven by pressures to succeed, fear of failure, and the temptation to cut corners. What has changed with the rise of GenAI is the sophistication of the tools available. Students can now use AI to generate essays, solve complex problems, and even pass exams with minimal effort. This video by Jason Tangen at the University of Queensland provides eye-opening examples of just how easily AI tools can handle complex assessment tasks.
Initially, many schools and universities responded to this threat by banning the use of GenAI tools. However, this approach has proven to be impractical. Students can easily bypass restrictions by accessing GenAI tools outside school networks, making enforcement almost impossible. More importantly, outright bans ignore the potential benefits that AI can bring to education. Instead of viewing GenAI as a tool for cheating, it is essential to see how it can be integrated responsibly to enhance learning.
The Potential of GenAI as a Learning Tool
GenAI offers significant educational benefits when used thoughtfully. It can provide personalised, real-time feedback, helping students to understand complex concepts, refine their ideas, and engage more deeply with the material. For example, students can use AI to draft responses, which they then revise and improve based on teacher feedback. This process can encourage them to think critically about their work, a skill that is essential for future success.
To leverage these benefits, teachers can adopt assessment strategies that make use of GenAI's strengths. Instead of designing assignments that require students to demonstrate memorisation or recall—tasks that AI can easily perform—consider assessments that emphasise skills at higher levels in Bloom’s taxonomy such as critical thinking, problem-solving, and the application of knowledge in new contexts. This can help students learn to use AI as a tool rather than a shortcut, promoting a deeper understanding of the subject matter.
Practical Strategies for Integrating GenAI
To navigate the integration of GenAI, educators can draw on frameworks that promote responsible and effective use. One example is the "ICE" model (Volante et al., 2023), which guides teachers in using AI-generated text as a starting point for deeper analysis and reflection through a series of three steps (Figure 1):
Figure 1. The ICE framework and Bloom’s taxonomy. From Volante et al., 2023.
Step 1: Understanding Ideas: Students fact-check key points in the AI content, developing research skills and ensuring a solid grasp of foundational concepts. Self- and peer-assessment activities help reinforce understanding and accuracy.
Step 2: Making Connections: Students improve the content by adding complexity and variety, making it less uniform. They connect ideas within the text and link them to personal experiences, fostering deeper analysis and engagement.
Step 3: Creating Extensions: Students extend the AI-generated ideas by introducing new perspectives, critiquing arguments, and applying concepts to real-world contexts. This promotes higher-order thinking and encourages authentic, creative learning.
By asking students to critique and improve upon AI outputs, this model encourages them to engage critically with content and develop higher-order thinking skills.
Another effective approach is to think of assessments as layers of Swiss cheese, which involves layering multiple types of evaluations (e.g., essays, oral exams, and practical tasks) at different times (Figure 2). Each assessment type has its own vulnerabilities (the holes in the cheese), but by combining them, the overall system is more robust (Rundle et al., 2020). This approach makes it harder for students to cheat consistently and ensures that, over time, assessments are measuring genuine understanding. The AI Assessment Scale (Perkins et al., 2024) can be a useful tool in this process, helping teachers to identify where and how AI tools can be usefully integrated into assessment design.
Figure 2. The ‘Swiss Cheese’ model for protecting academic integrity in assessments. Image by Ben Aveling (https://commons.wikimedia.org/w/index.php?curid=133912327) adapted from Rundle et al., 2020.
Shifting from Prevention to Preparation
Teachers play a crucial role in preparing students for a future where GenAI will be an integral part of everyday life and work. Instead of focusing solely on preventing misuse, educators can teach students how to use AI ethically and effectively. By integrating GenAI into learning, teachers can help students develop skills such as critical thinking, creativity, and digital literacy, which are vital for navigating an AI-enhanced world.
To support this shift, it is essential to have open conversations about how to use AI responsibly. Schools can establish clear guidelines on GenAI usage, provide training for both students and teachers, and create a culture that values ethical behaviour and transparency. This will help demystify GenAI, making it a tool that supports, rather than undermines, educational goals.
Ultimately, the key to integrating GenAI into education is balance. While it is important to address concerns about academic integrity, it is equally vital to recognise the opportunities that AI presents. When used thoughtfully, GenAI can enhance learning, streamline teaching, and prepare students for the demands of the modern world.
Dr Beth Chapman, Lecturer, Faculty of Education, University of Canberra
beth.chapman@canberra.edu.au
References
Perkins, M., Furze, L., Roe, J., & MacVaugh, J. (2024). The Artificial Intelligence Assessment Scale (AIAS): A Framework for Ethical Integration of Generative AI in Educational Assessment. Journal of University Teaching and Learning Practice, 21(06). https://doi.org/10.53761/q3azde36
Rundle, K., Curtis, G., & Clare, J. (2020). Why students choose not to cheat. In T. Bretag (Ed.), A Research Agenda for Academic Integrity. Edward Elgar Publishing. https://doi.org/10.4337/9781789903775.00014
Volante, L., DeLuca, C., & Klinger, D. A. (2023). Leveraging AI to enhance learning. Phi Delta Kappan, 105(1), 40–45. https://doi.org/10.1177/00317217231197475