The emergence of AI has become a defining issue for higher education worldwide, and Sweden is no exception. At the same time, emerging AI technologies reconfigure priorities and valuations within educational practice by mediating teaching and learning and opening new paths for how teachers and students relate to higher education practices. In this context, AI-mediated practices raise ethical questions that are often presented as unprecedented yet are deeply rooted in longstanding practices in higher education. This thesis undertakes an empirical exploration of AI-mediated practices in higher education, foregrounding teachers’ perspectives and focusing on the ethical issues arising from such mediations. Drawing on postphenomenology and technology mediation theory, the thesis examines how teachers perceive and experience emerging AI artefacts (automated grading systems (AGS) and generative AI (GAI) chatbots) in relation to their practices, and how these artefacts mediate their understandings of what they ought to do and how they ought to act when balancing sometimes competing demands of autonomy and accountability.

The thesis is a compilation of four complementary studies. Study I examines the ethical considerations of AGS, reviewing the literature on AGS and analysing their specificities through a relational ethics approach. This study highlighted that AGS not only introduce technical and procedural considerations but also reconfigure educational practices and relationships in ways that demand ongoing, situated, and relationally attuned ethical reflection. Study II is an interview study with AGS developers who are also university teachers using these systems. It examines their expectations, experiences, and the disruptions that AGS introduce into assessment practices. The findings underscore the ambivalent role of AGS as both promising and disruptive, offering efficiency and consistency, but also introducing ‘new’ frictions and ethical dilemmas. Study III is a study inspired by the Turing test, followed by focus group interviews with university teachers. It explores how GAI chatbots mediate teachers’ perceptions of their assessment practices. The findings indicate that the presence of GAI chatbots, allowing the possibility of AI-generated writing, shapes evaluation practices, prompting teachers to question authorship and, in some cases, reinforcing mistrust within the student–teacher relationship. Study IV is a focus group interview study examining how teachers experience and interpret the emergence of GAI and how it mediates their perceptions of their professional roles. Participants described GAI as both disruptive and potentially transformative. They were compelled to reconsider assessment formats, teaching priorities, and their responsibility to foster critical and ethical engagement with technology. 

The combined findings of the four studies show that the emergence of AI unsettles established practices and intensifies the uncertainties that characterise educational situations, placing greater demands on teachers’ professional judgment. The thesis also argues that the emergence of AI exposes and amplifies longstanding ethical issues, such as fairness, academic integrity, and equity, reshaping how these issues are understood and enacted as the technologies become embedded in higher education practices.

This paper examines how academic integrity is conceptualized and practiced in Swedish higher education in the context of generative Artificial Intelligence (AI) tools such as ChatGPT. It offers both an institutional and a student-centered perspective, drawing on university guidelines, pedagogical resources, and disciplinary cases, alongside a phenomenographic analysis of student reflections on AI use in academic work. The empirical data, gathered from 42 students in a course on academic writing, reveal a spectrum of attitudes—ranging from full transparency to pragmatic distinctions between substantial and auxiliary uses of AI. These reflections are situated within a broader cultural framework where academic integrity is treated not as a rigid code but as a relational and adaptive practice. The text argues that Sweden’s approach to academic integrity emphasizes trust, pedagogical support, and context-sensitive reasoning, rather than surveillance or prohibition. By analyzing institutional responses, cultural values, and student reasoning together, the article offers insights into how academic ethics are evolving in an era of AI-driven transformation.

Este artigo examina como a integridade acadêmica é conceituada e praticada no Ensino Superior sueco no contexto de ferramentas generativas de Inteligência Artificial (IA), como o ChatGPT. Oferece-se uma perspectiva institucional e centrada no aluno, com base em diretrizes universitárias, recursos pedagógicos e casos disciplinares, juntamente com uma análise fenomenográfica das reflexões dos alunos sobre o uso da IA no trabalho acadêmico. Os dados empíricos, coletados de 42 estudantes em um curso de escrita acadêmica, revelam um espectro de atitudes, que vai da transparência total a distinções pragmáticas entre usos substanciais e auxiliares da IA. Essas reflexões situam-se dentro de uma estrutura cultural mais ampla, na qual a integridade acadêmica é tratada não como um código rígido, mas como uma prática relacional e adaptativa. Argumenta-se que a abordagem sueca à integridade acadêmica enfatiza a confiança, o apoio pedagógico e o raciocínio sensível ao contexto, em vez de vigilância ou proibição. Ao analisar em conjunto as respostas institucionais, os valores culturais e o raciocínio dos alunos, o texto oferece reflexões sobre como a ética acadêmica está evoluindo em uma era de transformações impulsionadas pela IA.

Artificial intelligence (AI) chatbots trained on large language models are an example of generative AI which brings promises and threats to the higher education sector. In this study, we examine the emerging research area of AI chatbots in higher education (HE), focusing specifically on empirical studies conducted since the release of ChatGPT. Our review includes 23 research articles published between December 2022 and December 2023 exploring the use of AI chatbots in HE settings. We take a three-pronged approach to the empirical data. We first examine the state of the emerging field of AI chatbots in HE. Second, we identify the theories of learning used in the empirical studies on AI chatbots in HE. Third, we scrutinise the discourses of AI in HE framing the latest empirical work on AI chatbots. Our findings contribute to a better understanding of the eclectic state of the nascent research area of AI chatbots in HE, the lack of common conceptual groundings about human learning, and the presence of both dystopian and utopian discourses about the future role of AI chatbots in HE.

The emergence of generative artificial intelligence (GAI) has given rise to diverse narratives about its transformative potential in higher education. Despite widespread speculation about how GAI might change teaching and learning, there is a significant gap in understanding how GAI artefacts are perceived in educational practices, particularly from the perspective of university teachers. This study investigates how GAI mediates teachers’ practices and reconfigures professional roles. Drawing on post-phenomenology and technological mediation theory, we focus on university teachers’ experiences and perceptions of GAI in higher education. Twenty-four university teachers participated in workshops involving assessment exercises with GAI-generated outputs, followed by focus group interviews discussing the challenges and opportunities posed by GAI. Findings reveal that GAI prompts teachers to reassess established practices, particularly in relation to assessment, while confronting ethical concerns regarding fairness, trust, and quality. Teachers described their initial engagement with GAI as transformative yet challenging, as they navigated uncertainties about their roles while prioritising students’ learning and development. By capturing teachers’ voices during this pivotal period, the study contributes to the growing body of research on AI's role in higher education and provides a nuanced understanding of its impact on teaching and learning.

Automation in assessment is a fast-emerging AI research field that raises ethical issues for education. So far, dominant approaches to ethics have led to the development of numerous ethical guidelines to fix issues that the deployment of AI systems might introduce. This chapter critically examines the ethical considerations of AI automation in education by focusing on assessment and Automated Grading Systems (AGS). To this end, a relational approach to ethics is discussed that focuses on AGS’ specificities regarding data, algorithms, and assessment and the context where these systems are used, including situations and purposes, actors and relations, and time and place.

Interested in emerging technologies in higher education, we look at AI chatbots through the lens of human–technology mediations. We argue for shifting the focus from what higher education can do with AI chatbots to why AI chatbots are compelling for higher education’s raison-d’être. We call for a critical debate examining the power of AI chatbots in configuring students as civic actors in an increasingly complex and digitalized society. We welcome a continuous and rigorous examination of generative AI chatbots and their impact on teaching practices and student learning in higher education.

AI chatbots have recently fuelled debate regarding education practices in higher education institutions worldwide. Focusing on Generative AI and ChatGPT in particular, our study examines how AI chatbots impact university teachers' assessment practices, exploring teachers' perceptions about how ChatGPT performs in response to home examination prompts in undergraduate contexts. University teachers (n = 24) from four different departments in humanities and social sciences participated in Turing Test-inspired experiments, where they blindly assessed student and ChatGPT-written responses to home examination questions. Additionally, we conducted semi-structured interviews in focus groups with the same teachers examining their reflections about the quality of the texts they assessed. Regarding chatbot-generated texts, we found a passing rate range across the cohort (37.5 - 85.7%) and a chatbot-written suspicion range (14-23%). Regarding the student-written texts, we identified patterns of downgrading, suggesting that teachers were more critical when grading student-written texts. Drawing on post-phenomenology and mediation theory, we discuss AI chatbots as a potentially disruptive technology in higher education practices.

Den här handboken vänder sig till lärare som undervisar vid Stockholms universitet. Den är skriven med syftett att ge dig stöd för att aktivt motverka plagiering och fusk. Det finns även information om generativ AI.

This handbook, Preventing Plagiarism: A Guide for Stockholm University Staff (2024, 2nd Edition), provides a comprehensive framework to combat plagiarism and academic dishonesty in higher education. It addresses challenges arising from the increasing use of digital tools, including generative AI, and explores how these technologies impact student behavior and academic integrity. The guide emphasizes preventive strategies through effective teaching practices, transparent communication of expectations, and the design of assessments. It also includes legal insights, detection methods, and procedures for handling suspected plagiarism cases. By integrating practical exercises, tips for academic writing, and insights into ethical AI usage, this resource equips educators with tools to foster integrity and enhance learning outcomes.

As in other NLP tasks, Automatic Short Answer Grading (ASAG) systems have evolved from using rule-based and interpretable machine learning models to utilizing deep learning architectures to boost accuracy. Since proper feedback is critical to student assessment, explainability will be crucial for deploying ASAG in real-world applications. This paper proposes a framework to generate explainable outcomes for assessing question-answer pairs of a Data Mining course in a binary manner. Our framework utilizes a fine-tuned Transformer-based classifier and an explainability module using SHAP or Integrated Gradients to generate language explanations for each prediction. We assess the outcome of our framework by calculating accuracy-based metrics for classification performance. Furthermore, we evaluate the quality of the explanations by measuring their agreement with human-annotated justifications using Intersection-Over-Union at a token level to derive a plausibility score. Despite the relatively limited sample, results show that our framework derives explanations that are, to some degree, aligned with domain-expert judgment. Furthermore, both explainability methods perform similarly in their agreement with human-annotated explanations. A natural progression of our work is to analyze the use of our explainable ASAG framework on a larger sample to determine the feasibility of implementing a pilot study in a real-world setting.