Artificial Intelligence in Education: From Pedagogical Decisions to Social Consequences

Artificial intelligence is becoming a staple of modern education and is increasingly influencing the ways in which we teach, learn, and assess knowledge. Its use raises numerous questions that go beyond technology: how to design meaningful learning tasks, how to maintain trust in academic practices, and how to prepare students for a world in which collaboration with algorithms will be an everyday reality.

The debate on artificial intelligence (AI) in education is increasingly moving away from simple bans and control measures and is shifting toward the question of how to design teaching and assessment practices that take into account the reality of generative AI. The authors emphasize that AI exposes the weaknesses of traditional knowledge assessment, which is based primarily on reproduction or final products without insight into the thought process (more in the article AI-resilient and creative HE assessment design). Consequently, tasks that require reflection, justification of decisions, and the integration of knowledge with personal experiences are coming to the forefront. To assist teachers in this process, a framework of five key stages is presented, leading from the clear definition of learning objectives, through task design and the use of AI, to the validation and reflection on the learning process (more in the article A practical tool for valid assessments in the GenAI-enabled university). This approach does not prohibit the use of AI, but rather systematically integrates it into the learning environment; the pedagogical challenge is not technical in nature, but remains primarily one of content.

Research from the University of Westminster in London shows that students most often use generative artificial intelligence to explain unclear concepts, check their understanding, develop ideas, structure assignments, improve the language of their texts, and summarize materials; they use it significantly less often to verify facts or sources (more in the article What your students are actually doing with GenAI). Students generally view AI as a learning aid or study partner, not as a substitute for their own thinking, and it proves particularly useful for international students and those facing language challenges. Students also emphasize that they want universities to provide clear, specific, and subject-specific guidelines on what is permitted for each assignment and how to properly cite the use of AI. They also consider it important for educational institutions to offer support in learning how to formulate effective prompts and in developing critical judgment and verifying answers (more in the article Students told us what GenAI guidance works). This is crucial for education, as thoughtful and guided use of AI can enhance learning, critical thinking, and fairness, while unclear rules and a lack of guidance increase uncertainty and pose risks to academic integrity.

Despite the rapid development of AI, there remain areas where human judgment is indispensable. An important caveat in discussions about the use of artificial intelligence in academia is that AI cannot take on the role of an equivalent expert evaluator. This is particularly evident in peer-review processes, where AI cannot replace the contextual understanding, accountability, and ethical judgment that are inextricably linked to human expertise. The same applies to research work, where a division of labor makes sense: AI can effectively support routine and time-consuming tasks, while humans retain control over interpretation, evaluation, and final decisions (more in the article Campus talks: ‘Use your brain!’ And other pointers from a seasoned computer scientist on using AI in research). Such a balance is also crucial in education, where AI must remain a tool to support learning, not an authority or a substitute for understanding. Indeed, pedagogical value arises precisely from the distinction between the automation of processes and the development of professional judgment, which is shaped through experience, mentorship, and reflection (more in the article Balance human intuition with machine efficiency in scientific research).

Among the latest developments in the field of AI, new initiatives from OpenAI stand out, pointing the way toward the development of tools for the educational sector as well. The unveiling of the Codex app raises questions about the future of programming and computer science education, where understanding code is becoming more important than writing it by hand (more in the article Introducing the Codex App). The OpenAI Frontier initiative and the Prism tool, meanwhile, demonstrate a broader interest in developing advanced models and gaining a better understanding of how they work and the impact they have (more in the articles Introducing OpenAI Frontier and Introducing Prism). For education, this means there is an additional need to explain what these tools can and cannot do, and how to use them critically. Technological progress thus directly influences learning objectives and curriculum decisions.

At the institutional level, discussions are focused on risk management issues and the long-term implications of introducing artificial intelligence into higher education (for more, see the articles Managing AI Risk and Return and When AI Meets Data: The Promise and the Pressure of Bringing AI into Higher Education Systems). The issue of embedded digital accessibility is becoming particularly important, as there is a risk that new tools will exacerbate existing inequalities (see the article The Case for Embedded Digital Accessibility). At a conference on artificial intelligence in education, organized by the National Council’s Commission for Education, Culture, Science, Sports, and Youth and the Development Council of the Slovenian Academy of Sciences and Arts (SAZU), participants emphasized that the introduction of artificial intelligence into the education system is essential but requires a thoughtful and systematic approach, including teacher training—a point also confirmed by broader international discussions on the institutional role in AI governance (for more, see the articles At the Roundtable, a Majority Supports Introducing Artificial Intelligence into the Education System and Roundtable: Artificial Intelligence in Education). Institutions play a key role in establishing clear frameworks that extend beyond individual subjects or teachers. AI thus becomes a matter of educational policy, not merely pedagogical practice.

The development of artificial intelligence is increasingly influencing digital environments where communication, learning, and social interaction take place. The articles highlight the emergence of new social networks designed primarily for the operation of AI agents, where a significant portion of the content is created and shared by artificial agents, while humans enter such environments primarily as observers or participants who adapt to the logic of AI operation (more in the articles I hung out inside Moltbook, the AI-only social network where humans are the outsiders and Humans are infiltrating the social network for AI bots). Such examples raise questions about the authenticity of participation, the distinction between human and artificial interlocutors, and what learning even means in environments where it is unclear who is the author of knowledge. At the same time, public debate is intensifying about the growing role of chatbots in everyday life, work, and education, and about how quickly they are becoming part of decision-making and opinion-forming (more in the article Battle of the chatbots: Anthropic and OpenAI go head-to-head over ads in their AI products). Education therefore plays a crucial role in developing critical literacy, which enables individuals to understand these environments, recognize their limitations, and participate responsibly in digital society.

To wrap things up, let’s shift our focus slightly away from education and turn to current events—the Winter Olympics—which serve as a significant example of the visible and extensive use of artificial intelligence in practice. AI will influence how competitions are followed, data is analyzed, viewing experiences are personalized, and even support referees’ decisions (for more, see the articles The Technologies Changing How You’ll Watch the 2026 Winter Olympic Games, From chatbots to replays: Alibaba rolls out AI suite for 2026 Winter Olympics, and AI is coming to Olympic judging: what makes it a game changer?). These shifts raise questions of transparency, trust, and accountability that are surprisingly similar to those in education. The Olympic Games thus offer a concrete example that teachers can use to introduce students to discussions about AI outside the academic setting. It is an opportunity to connect learning with current events.

Artificial intelligence in education clearly demonstrates that technological progress alone does not lead to higher-quality learning. The key question is how educational institutions, teachers, and students make responsible decisions about its use. The newsletter highlights the need for thoughtful pedagogical approaches, clear guidelines, and the development of competencies that enable critical judgment and adaptation to change. Thus, ICT serves as a turning point that determines whether education will merely follow technology or will be able to meaningfully integrate it to support learning.

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