Abstract:

In recent years, generative artificial intelligence (GenAI) has been progressively incorporated into the university environment as a support tool for a wide range of academic activities. These tools can assist students in tasks such as searching for information, synthesizing or generating texts, and supporting problem solving. As a result, if used properly, they could enable students to approach complex assignments more efficiently, while accessing a wider range of resources.

However, the growing presence of GenAI in higher education has also raised important pedagogical, ethical and methodological challenges. Particularly, regarding the development of critical thinking, academic integrity and the responsible use of these tools.

The objective of this work is to promote a critical perspective on the use of GenAI when interpreting financial information and making business decisions. It aims to encourage students to remain in control of the process, identifying what information is necessary and how it should be incorporated into prompts, so that interpretations are accurate and meaningful.

To promote the responsible and pedagogically meaningful use of GenAI, a team of professors from the Universitat de Barcelona designed a three-phase activity for the first-year Business Economics course in the Bachelor’s Degree in Business Administration and Management. The activity was carried out over several weeks during the first semester of the 2025–26 academic year within the topic “Sources of Business Financing” and involved nearly 400 students from five class groups.

- Phase 1 (individual and remote). Students were provided with balance sheet data, from a random company and asked to answer a set of economic-financial questions using an AI assistant, at their choice. Here, the AI acted as the main agent in generating responses.
- Phase 2 (individual and remote). Original data, in tandem with additional contextual information, was made available, requiring students to reconsider previous choices. Here, students answered the same questions and were expected to act as the main agent: guiding the GenAI assistant responses based on the new insights.
- Phase 3 (synchronous).

The activity concluded with a collective class discussion, and a follow-up survey asking students for their reflections on:
(i) the activity itself and
(ii) the potential environmental impact of the use of GenAI.

Data was collected from students’ questionnaire responses (3 phases) and their AI prompts (phases 1 and 2). As expected, GenAI in the first phase produced responses that appeared coherent but were ultimately incorrect, as information was incomplete. In the second phase, results aligned very much with the correct solution, as more reasoning was incorporated into the prompts.

The results highlight some relevant findings. First, students changed answers between phases 1 and 2, showing the importance of contextual (complete) information in financial analysis, as AI-generated responses may often led to convincing but inaccurate interpretations. Second, preliminary analysis of reflections in phase 3 evidence that, broadly, students understood the importance of being critical and context-aware when guiding GenAI. Also, the impact of AI on environment seems still unclear among students. Overall, the experience suggests that carefully designed pedagogical activities can transform AI from a shortcut for completing tasks into a valuable tool for fostering critical thinking in higher education

Keywords:

AI in Higher Education, critical thinking, AI sustainability impact.