Artificial intelligence challenges us, inspires us, and guides us toward change

Since we just mentioned exam season—in the United Kingdom alone, nearly 7,000 cases of student cheating involving artificial intelligence were recorded last year, a significantly higher number than the previous year. And experts warn that this figure is most likely just the tip of the iceberg (source). Despite this trend of cheating, many universities are turning a blind eye to reality, as acknowledging this problem usually involves various costs; however, avoiding and failing to address this challenge can cause far greater damage in the long run (source).

At our university as well, we witnessed a case of group cheating by students during exams in the Moodle virtual learning environment last week, which occurred despite the exam being proctored. Using a browser extension (Crowdly.ai), the students collectively solved Moodle quizzes and gained insight into the correct answers to more than 200 different questions.

This case—and other similar ones around the world, of which there are, judging by what has been reported so far, a great many—can be understood not only as a threat but also as an opportunity. Although the rise in cheating among students has led us in many cases to revert to traditional exams, where students write on paper (source), we are now at a time when we can consider opportunities to change assessment methods so that they actually test knowledge rather than merely the end product. As with any “crisis,” this one also presents opportunities—a theme that was central to the Digital Universities US conference, where participants took a hard look at somewhat outdated assessment models and emphasized the need to foster greater trust in education, transparency, and the reform of the educational system (source).

There are many ways to approach this.

To begin with, we thought it would be appropriate to highlight a case where German educators in teacher-training programs used large language models to provide feedback to students during exams. The results of the study showed that students were satisfied with the AI-generated feedback, and its use had a positive impact on their subsequent writing (source).

Australia has taken this a step further by developing a AI model that analyzes five different aspects of student engagement upon enrollment and predicts the risk of dropping out. This enables them to provide timely support to students, including those whose grades do not otherwise indicate a risk of dropping out but who are found to have weak personal and professional motivation for their studies (source).

And it is precisely such stories or use cases that can also lead to greater student proactivity. We would like to highlight the example of a student at Mercer University who developed a curriculum for elementary and high school students that includes reflection on how AI works, image recognition games, creating filters, and reflection on ethical use. After the initial implementation of the curriculum, it became clear that this approach can encourage even less-motivated students to participate (source).

We also found the comprehensive example of educational reform at Babson College interesting, where they opted for a proactive approach: they developed a certified program for students to learn about AI, provided MS 365 Copilot licenses for all staff and students, developed a math tutor to assist students struggling in this area, use an AI dashboard to monitor AI usage and participation in funded activities in this field on a weekly basis, they have hired an AI analyst and established an AI lab, where they have opportunities for discussion and training in this field, as well as a support program for teachers, where they experiment with AI and plan its use in the pedagogical process (source).

Although there are several ways to approach leveraging the benefits of AI in education, a challenge remains for educators who do not feel confident enough to use AI and therefore need support (source). We can also infer this from the significant lack of research on the use of AI for teachers’ professional development (source). Consequently, various support institutions are addressing this issue in different ways. We found the example of support from the British Ministry particularly interesting; although it addresses a lower level of the educational hierarchy, it nevertheless offers a range of ideas and support for developing reflections on AI in education, including at the systemic level (source).

That is why we invite you: to help shape the future:

• try out various AI tools,
• reflect on ethical dilemmas,
• create and/or update existing AI-based learning activities, and
• develop your own strategies for responsibly integrating AI into the classroom.

Ask yourselves whether AI can help you provide feedback, whether you can maintain academic integrity using existing methods of assessment and evaluation, and how you are preparing students for a future that—in our opinion—almost inevitably involves skills for working with AI.

We hope that at least some of the news items presented here will provide you with at least a glimpse of the kind of education we envision in the age of artificial intelligence.

Every story in today’s newsletter could also be your own. We just have to decide where to start.

To review older news items and stories, please visit our resources.

Authors: Sanja Jedrinović, Mateja Bevčič, Eva Kern Nanut, Eva Škraba, University of Ljubljana Center for the use of ICT in pedagogical process