The increasing use of generative AI in academic writing has created new challenges for academic integrity, including AI-based plagiarism (AI-giarism). This study assessed university students’ knowledge, attitudes, and practices (KAP) regarding AI-giarism in Bangladesh. A cross-sectional quantitative survey was conducted among 386 students from public and private universities in Bangladesh using a structured questionnaire. Data were analyzed using descriptive statistics, independent-samples t-tests, one-way ANOVA, and multiple linear regression. Students demonstrated moderate knowledge and mixed practices regarding AI-giarism, while their attitudes were generally favorable toward academic integrity. No significant differences in KAP were observed by gender, age, education level, or research experience. However, attitudes differed significantly by university type, residential status, publication experience, and academic discipline, whereas academic writing training was associated with better-reported practices. Students with greater familiarity with AI demonstrated higher knowledge. In the regression analyses, AI familiarity, attitude, and practice were significantly associated with knowledge, whereas residential status and publication experience were associated with attitude. Knowledge was the only significant variable associated with practice. The findings indicate gaps between students’ understanding and responsible AI-related academic practices. The results highlight the need for clear institutional guidance, academic integrity and AI literacy training, and assessment practices that promote transparent and responsible use of generative AI.
Much of the ethical debate about artificial intelligence turns on a single question: do AI systems behave in line with human preferences, norms, or regulation? This question has primarily been the focus in AI ethics. This paper offers a conceptual and philosophical contribution. Here we give an alternative basis for AI ethics by evaluating the role technology plays in building and stabilising worlds of meaning. Experience is modelled as passing through nested representational layers: the world 𝒲 , the screen of perceived reality 𝒲' on which agents encounter and interpret the world, and the artificial screens 𝒲” built by digital platforms, recommender systems, and immersive environments. Meaning is formed, held, and acted on within a minimal cycle of perception, integration, and action. On that basis, many familiar ethical problems around AI show up less as simple failures of behaviour or compliance and more as breakdowns in semantic coherence, that is, in the preservation of meaning along the chain 𝒲→𝒲'→𝒲” . Perceptual sovereignty complements semantic coherence by naming agents’ capacity to understand, contest, redirect, or exit such mediation. Fairness, transparency, accountability, privacy, and safety are recast as aspects of these deeper conditions of coherence and agency, not as free-standing technical or policy goals. We derive a four-level hierarchy of ethical depth (L _0 –L _3 ) and sketch governance aimed at the integrity of constructed worlds, with consequences for how risk, responsibility, and regulation are conceived when technological mediation is everywhere.
Wikipedia is the most widely used reference work in history and one of the most valuable training resources available to the generative AI (genAI) industry; and yet that same industry is systematically undermining the conditions that make Wikipedia possible. GenAI companies extract Wikipedia’s corpus to train their products while redirecting users away from the encyclopedia. The result is declining readership, a shrinking volunteer base, and a growing threat to the encyclopedia’s long-term sustainability as a knowledge commons. This article examines that threat through the lens of commons theory, arguing that Wikipedia functions as a knowledge commons that is collaboratively constructed, transparently produced, and freely accessible, and that genAI-based tools operate as a force against it. Against this backdrop, the article makes the case that academics have a responsibility to defend Wikipedia and address the epistemic injustice caused by the genAI industry’s attempts to consolidate control over knowledge production.
The effectiveness and the moral credibility of the European Union’s Artificial Intelligence Act will be decided less by its substantive provisions than by the national institutions charged with enforcing them. This article examines Germany’s KI-Marktüberwachungs- und Innovationsförderungsgesetz (KI-MIG), adopted by the Bundestag on 11 June 2026 and in force since 29 July 2026, as one of the first comprehensive national enforcement frameworks for the AI Act and as a test case for the claim that institutional design is itself an ethical variable. The article reconstructs the KI-MIG’s architecture—the centralisation of market surveillance at the Federal Network Agency, a coordination and competence centre (KoKIVO), an independent market surveillance chamber for fundamental-rights-sensitive systems, a central complaints body, and a hybrid model that preserves sectoral competences—and evaluates it against five criteria of legitimate enforcement derived from regulatory theory: independence and mandate fidelity, competence and capacity, accessibility and voice, coherence, and regulatory candour. The assessment yields verdicts rather than a mere inventory. The design earns genuine credit on coherence and on accessibility, but the article concludes that the independence construction for fundamental-rights-sensitive supervision—a chamber staffed, in personal union, by the leadership of a hierarchically embedded economic regulator—falls short of the AI Act’s heightened requirements. Further concerns—dispersed notification responsibilities, undersecured Länder capacity, and an official rhetoric of ‘lean oversight’ that risks normalising under-enforcement—are sharpened by the Digital Omnibus’s deferral of high-risk obligations to December 2027. The article concludes with transferable design principles for jurisdictions that are now building AI enforcement institutions.
In debates on artificial intelligence in healthcare, trust is frequently invoked as an ethical requirement and equated with technical properties such as transparency, explainability, and performance. This paper distinguishes technical reliability from trust in AI-supported clinical decision-making, arguing that the two belong to distinct registers that current debates tend to elide. The argument is conceptual and case-based, drawing on the European MIRACLE project, which develops a machine learning algorithm to support adjuvant therapy decisions in resected early-stage non-small-cell lung cancer. The paper first examines how technical reliability is conceptualized during algorithm development, then argues that reliability, though necessary, cannot by itself generate trust. Drawing on the bioethical literature on trust in the therapeutic relationship, it holds that trust is relational, involves vulnerability and responsibility, and cannot be attributed to artificial intelligence systems as such: its conditions lie in the relational and organizational arrangements within which algorithms are deployed. On this basis, the paper proposes replacing the expression "trustworthy AI" with "technically reliable AI," a distinction that relocates the conditions of trust from the design of the system to the clinical settings in which it is used.
