Academic League of Artificial Intelligence: An Integrative Perspective of Teaching, Research, and Extension

Synthesis: Panisson et al. (2026) present the organizational framework of the Academic League of Artificial Intelligence (LIA) at the Federal University of Santa Catarina (UFSC), a student-centered, project-based structure that integrates teaching, research, and university extension. The framework combines democratic governance, Collaborative Learning, and dynamic project organization to develop both technical and transversal competencies. Representative initiatives include competition teams, study groups, open lectures, knowledge repositories, and AI-powered applications with social impact. The paper argues this is a flexible, replicable model for embedding AI in Education and AI Literacy into engineering and computing education through extracurricular, community-connected structures rather than formal coursework alone.

Key Findings

1. LIA at UFSC demonstrates an organizational framework that integrates teaching, research, and university extension through a student-centered, project-based approach to AI in Education, offering a documented model that was previously absent from the computer science and engineering literature.
2. Democratic AI Governance—an annually elected student administrative board supervised by faculty mentors—ensures institutional continuity across generations, while a dynamic portfolio of projects keeps the league adaptable to emerging AI topics and student interests.
3. A single structure supports diverse outcomes: competition teams, peer-led study groups, open lecture series, knowledge repositories, and AI-powered applications with social impact each yield educational, scientific, and extension products.
4. The framework is a flexible, replicable reference model for embedding AI Literacy and community engagement into engineering and computing education, offering practical guidance for academic leagues and similar student organizations.

A Model for Extracurricular AI Education

Academic leagues are mechanisms for promoting extracurricular education and strengthening university-society integration. The Academic League of AI at UFSC demonstrates how a Collaborative Learning community can organize around AI education without being bound to a formal curriculum. Its framework rests on three pillars of the university mission—teaching, research, and extension—coordinated through democratic student governance. Rooted in Brazil's university system, the model exemplifies how AI Literacy can spread through interest-driven, extracurricular structures rather than formal coursework alone, and it functions as a Community of Inquiry in which students, faculty mentors, and the broader community engage around artificial intelligence.

Democratic Governance and Organizational Structure

LIA pairs a stable administrative structure with a dynamic portfolio of projects. A student-elected administrative board—composed of a President, Vice-President, Secretary, and Communication Coordinator—is chosen annually through a democratic process, while faculty members act primarily as advisors and mentors. This AI Governance model allows students to assume the leading role in the league's strategic direction while preserving continuity across successive generations. Projects, in contrast, remain flexible: each is coordinated by a student leader elected by its members, and weekly meetings integrate the activities of all active projects. The separation between administrative management and project execution has proven essential to the league's Sustainability, fostering student leadership, protagonism, and collaborative decision-making.

Learning Through Project Organization

Projects such as competition teams and open lectures create Project-Based Learning opportunities that develop leadership, scientific production, and community engagement. Study groups form the foundational Collaborative Learning mechanism, letting students investigate machine learning, computer vision, large language models (Large Language Models (LLMs)), and multi-agent systems before engaging in research or development. Knowledge repositories institutionalize what members learn, while AI-powered applications with social impact connect Computational Thinking to real civic problems, embodying Active Learning, Self-Directed Learning, and AI Literacy in practice.

Representative Projects

Representative initiatives illustrate how a single organizational structure produces educational, scientific, and extension outcomes. The LI(A)RA competition team developed intelligent agents for the Multi-Agent Programming Contest, integrated with an undergraduate course, and transformed its results into a booklet and a self-instructional online course. The "Open Talks about Artificial Intelligence" cycle promotes dialogue between the university and society, with recorded lectures made publicly available. The Repository of Knowledge in Artificial Intelligence (RepoAI) consolidates tutorials and implementation examples as an Open Source knowledge base supporting member onboarding and knowledge preservation. Study groups underpin every initiative, and the GAIA project develops AI-powered interactive applications with social impact—including Forca Libras—addressing education and Accessibility challenges through serious games and rehabilitation technologies, promoting digital inclusion and democratizing access to AI knowledge.

Replicability for AIED

The reported experience offers practical guidance for other academic leagues and student organizations seeking to integrate the three university pillars. It complements formal AIED instruction by providing informal, interest-driven pathways into AI in Education, and it illustrates how transversal competencies such as Critical Thinking, leadership, and Lifelong Learning can be fostered through Student Experience-centered AI Governance rather than top-down curricula. The framework demonstrates an approach to Curriculum Design and engineering and computing education that scales across computer science programs and extends beyond Brazil's university context.

What this means for practice

• Learners. Run the organization, not just the projects: LIA's annually elected student board (president, vice-president, secretary, communication coordinator) and project leaders elected by their own members let students set strategic direction while faculty serve as mentors.
• Learners. Turn what you build into persistent artifacts — tutorials, repositories, technical reports, recorded lectures, online courses — because that is what lets later cohorts extend the work instead of recreating it.
• Administrators. Start a league without new infrastructure: none of LIA's representative projects required large financial investment or specialized hardware, with success depending on student engagement, faculty mentorship, and collaborative organization.
• Instructors. Take the mentor role and keep the governance separate from project execution: weekly meetings integrate all active projects and annual elections preserve continuity, and that separation is what sustains student leadership.
• Researchers. Document organizational models and then measure them: the paper's own gap statement is that computer science and engineering lacked documented models of this kind, and it calls for multi-institution studies with quantitative learning-outcome measures.

Limitations

• Single case: the reported experience comes from one academic league at one institution — LIA at the Federal University of Santa Catarina — a limitation the authors state directly.
• The evaluation is primarily qualitative and focused on the framework's organizational aspects; the paper reports no quantitative assessment of student learning, retention, leadership development, or engagement, and names those measures as future work.
• Replicability beyond the founding context is argued rather than demonstrated: that the organizational principles are largely independent of the technical domain, and that the model extends beyond Brazil's university context, is asserted without a second instantiation.
• No data-collection procedure is described — no participant counts, surveys, or interviews — so the account of representative projects rests on the organizers' own description rather than on independent observation or participant data.

Citation

Panisson, A. R., Vianna, M. E. W. M., da Silva, I. F., da Silva, H. H., Savaris, R. F., Costa, B. P., et al. (2026). Academic League of Artificial Intelligence: An Integrative Perspective of Teaching, Research, and Extension.