AI & Society最新文献

{"title":"Ethics and the meaning of life: the boundaries of artificial intelligence","authors":"Sigifredo Laengle,&nbsp;Leonardo Torres","doi":"10.1007/s00146-026-03119-1","DOIUrl":"10.1007/s00146-026-03119-1","url":null,"abstract":"<div><p>Can a machine become autonomous, operating outside human control and bearing responsibility for its actions? These questions require us to clarify what fundamentally distinguishes humans from machines with respect to responsibility. This article analyzes the current philosophical debate through phenomenological perspectives, focusing on the thought of M. Henry, E. Levinas, and V. Frankl. We find that current discourse emphasizes AI’s functional limits, whereas our research identifies essential ontological differences. Specifically, we argue that humans possess an exclusive, irreducible drive to seek meaning, while machines lack this capacity. For machines, ethics is reduced to algorithmic execution within a distributed network. Ultimately, when AI acts unpredictably or harmfully, it is because only humans can assume—and fail to assume—moral responsibility. We conclude by proposing the search for meaning as the ontological boundary and substantive criterion for responsibility, thereby contributing to the ongoing philosophical debate.</p></div>","PeriodicalId":47165,"journal":{"name":"AI & Society","volume":"41 9","pages":"8349 - 8358"},"PeriodicalIF":6.1,"publicationDate":"2026-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149098938","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The intelligence we need now","authors":"Douglas Schuler","doi":"10.1007/s00146-026-03152-0","DOIUrl":"10.1007/s00146-026-03152-0","url":null,"abstract":"","PeriodicalId":47165,"journal":{"name":"AI & Society","volume":"41 6","pages":"5461 - 5465"},"PeriodicalIF":6.1,"publicationDate":"2026-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148433434","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The deskilling dilemma: ChatGPT, pedagogical obligation, and the paradox of journalism education","authors":"Md. Sazzad Hossain,&nbsp;David Dowling","doi":"10.1007/s00146-026-03139-x","DOIUrl":"10.1007/s00146-026-03139-x","url":null,"abstract":"<div><p>This paper critically examines the perceived impact of ChatGPT on journalism and mass communication education in the United States, with particular attention to how the technology is framed by instructors within the field. Using a survey of 91 educators, the study explores perceived threats, benefits, and pedagogical adaptations related to generative AI. Drawing on critical perspectives from science and technology studies, the paper interrogates the notion of “disruption” in journalism education, analyzing how AI tools are reshaping assumptions about authorship, expertise, and the future of media labor. While some educators see promise in using ChatGPT for writing assistance or curriculum augmentation, others express concern over misinformation, plagiarism, and the deskilling of critical journalistic judgment. This paradoxical tension exists between acknowledgment of the responsibility to teach proper use of ChatGPT, given its central role in current journalistic production practice, and acute awareness of how overreliance on generative AI risks not only deskilling but also eroding the expressive and epistemic labor central to journalism. Based on Giroux’s critical pedagogy and Christensen’s theory of disruptive innovation, the study calls for a reflexive, ethically grounded approach to integrating generative AI, informed by a broader understanding of technological governance and academic labor in the digital university.</p></div>","PeriodicalId":47165,"journal":{"name":"AI & Society","volume":"41 9","pages":"8461 - 8479"},"PeriodicalIF":6.1,"publicationDate":"2026-06-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s00146-026-03139-x.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149098975","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Generative AI and generativity in education: implications of temporality and succession in human life","authors":"Albrecht Fritzsche","doi":"10.1007/s00146-026-03143-1","DOIUrl":"10.1007/s00146-026-03143-1","url":null,"abstract":"<div><p>Generativity has recently gained attention as a property of a specific class of algorithms in machine learning. However, the term has also been used elsewhere with a different meaning. In particular, it plays an important role in Erikson’s model of human development, with many implications for education from psychological and sociological points of view. For research on applications of generative artificial intelligence (AI) in education, these points of view are relevant as well. They allow for a comparative investigation of different views on generativity, which are described here as psychosocial views on the one hand and technical views on the other. What psychosocial views add to research on generative AI is a reference to temporality and succession as key issues in human life. Both play an important role in education. Furthermore, they cause tensions in the transition from student to teacher, adding dialectic dynamics to education and the overall relationship between generations. Applications of generative AI in education leave little space for these dynamics to be acted out between humans and their human successors. At the same time, tensions become visible elsewhere, regarding humanity as such and its succession. Considering all this, new questions must be asked about the function of education in society and its ability to mediate transitions between generations.</p></div>","PeriodicalId":47165,"journal":{"name":"AI & Society","volume":"41 9","pages":"8197 - 8206"},"PeriodicalIF":6.1,"publicationDate":"2026-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149099155","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Language and time in education: thinking in the age of generative artificial intelligence","authors":"Shengmeng Cui","doi":"10.1007/s00146-026-03148-w","DOIUrl":"10.1007/s00146-026-03148-w","url":null,"abstract":"<div><p>In today’s world of Generative AI (GenAI), the challenges in education are often reduced to simple questions: ‘How do we stop cheating?’ or ‘How can we work faster?’ However, this paper poses a deeper question: When machines can speak and write for us, what does it mean for a human to think? Drawing on years of experience teaching English, the author observes that schooling often transforms learning into calculative thinking—a relentless pursuit of standard answers, grades, and future job prospects. This trend encourages students to treat language as a warehouse of information that can be easily outsourced to AI. Borrowing from the philosopher Martin Heidegger, this paper argues that thinking is not just about producing an output; it is a way of being in the world and ‘dwelling’ within language. The article explores how modern education distorts our sense of time, causes our judgment to wither into mere calculation, and allows GenAI to fragment our language. It reveals how technology subtly pressures us to become as efficient and predictable as the machines we use. Finally, the author proposes a new educational stance called answerability without exhaustiveness. This approach calls for ‘meditative thinking,’ reimagining education as a journey to discover our connection to the world, rather than a factory for producing compliant text.</p></div>","PeriodicalId":47165,"journal":{"name":"AI & Society","volume":"41 9","pages":"8221 - 8229"},"PeriodicalIF":6.1,"publicationDate":"2026-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s00146-026-03148-w.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149099154","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Responding to the normativity problem in machine ethics through the Power of the Other","authors":"Yuhang Guo","doi":"10.1007/s00146-026-03131-5","DOIUrl":"10.1007/s00146-026-03131-5","url":null,"abstract":"<div><p>Contemporary debates in the source of normativity in machine ethics display a growing tendency to separate instrumental rationality from practical reason. The rationalist approach understands AI as an extension of the ideal rational agent, presupposing strong autonomy to ground normativity in internal coherence. Rationalist approaches exclude the possibility of understanding the external world through instrumental means and therefore fail to provide an adequate answer to the problem of normativity. The reasons why individuals act morally are not reducible to the validity of rational argumentation itself. Rather, any response to the problem of normativity must be grounded in respect for humanity and in the recognition of one’s own intrinsic value. The externalist approach treats machines as bearers or executors of moral law and can be divided into rigorist and teleological models. Yet, in both cases, AI is ultimately confined to an instrumental existence, and moral responsibility becomes difficult to sustain insofar as human agents are characterized by weak autonomy or problematized forms of autonomy. By invoking the Power of the Other in response to structural pressures in digital life, rather than compelling individuals to conform to algorithmic logic, this perspective promotes an open moral framework. It emphasizes ethical engagement grounded in relational awareness, enabling individuals to navigate the complexities of algorithmically mediated environments while maintaining attentiveness to the Other and the ethical significance of their actions.</p></div>","PeriodicalId":47165,"journal":{"name":"AI & Society","volume":"41 on","pages":"7189 - 7204"},"PeriodicalIF":6.1,"publicationDate":"2026-06-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148807565","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Out of time for the version of record: designing living knowledge infrastructure for AI-accelerated scholarship","authors":"Hussein AlAhmad","doi":"10.1007/s00146-026-03126-2","DOIUrl":"10.1007/s00146-026-03126-2","url":null,"abstract":"<div><p>Dominant publishing workflows took shape under conditions of relative epistemic stability: slower problem cycles, bounded corpora, and institutions better able to absorb delay. Those conditions now hold unevenly, and in many fast-moving fields they no longer match the tempo of the object itself. Claims can become decision-relevant within days or weeks, while formal validation still unfolds over months. The result is a widening gap between what begins to circulate with authority and what has actually been epistemically stabilised. This article treats generative AI not as the origin of that problem, nor as a substitute for scholarly judgement, but as an intensifier of pressures already building within contemporary systems of validation. Using conceptual-design research, it identifies recurring breakdown patterns in scholarly publishing and translates them into explicit design requirements. From there, it proposes Living Knowledge Infrastructure (LKI) as a redesign of scholarly validation for an era in which generative AI accelerates knowledge production faster than existing publishing systems can verify it. It advances a sociotechnical framework for sustaining cumulative verification beyond the version of record. The contribution is architectural and testable: LKI links pedagogical review, executable knowledge, journals as knowledge stewards, and epistemic care into a model designed to preserve review continuity, lower verification costs, and strengthen corrigibility under accelerated conditions. Rather than calling simply for faster or more open dissemination, the article argues for a linked redesign of how scholarly judgement is recorded, carried across versions, and made publicly accountable when uptake begins to outrun validation.</p></div>","PeriodicalId":47165,"journal":{"name":"AI & Society","volume":"41 9","pages":"8393 - 8405"},"PeriodicalIF":6.1,"publicationDate":"2026-06-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149098590","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AI and extended authenticity: autism as a case study","authors":"Joel Krueger","doi":"10.1007/s00146-026-03101-x","DOIUrl":"10.1007/s00146-026-03101-x","url":null,"abstract":"<div><p>I consider some ways autistic people use AI chatbots for companionship and connection, and some distinctive benefits and harms that may follow from this usage—particularly as it relates to the virtue of authenticity. Drawing on first-person reports, as well as debates about the extended mind and extended virtues, I argue that while AI systems lack authenticity, they may be brought into extended (human-AI) systems that help some users achieve authenticity. AI companions may prompt some autistic persons to realize authenticity-expressing traits and capacities that might not otherwise emerge. Along the way, I consider some important ethical concerns. I conclude that while we should remain critical in assessing the place of AI in our lives more generally—and the motives behind Big Tech’s relentless promotion of it—we also should not be overly hasty in dismissing these digital connections as necessarily inauthentic or lacking value.</p></div>","PeriodicalId":47165,"journal":{"name":"AI & Society","volume":"41 9","pages":"8231 - 8246"},"PeriodicalIF":6.1,"publicationDate":"2026-06-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s00146-026-03101-x.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149098592","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Resisting AI: a conceptual framework for practices of resistance against artificial intelligence in the creative industries","authors":"Daniel Vasconcelos","doi":"10.1007/s00146-026-03135-1","DOIUrl":"10.1007/s00146-026-03135-1","url":null,"abstract":"<div><p>We have witnessed a burgeoning debate on the nature of AI creativity and its implications for the creative industries, particularly in relation to labour dynamics. However, even though we can find critical studies on AI and creative work, more research is needed to investigate, empirically or conceptually, the sources, methods, and forms of resistance against AI in the creative industries. By pivoting the reference of analysis to concentrate on the agency of workers in accepting, resisting, and navigating these technologies, this paper brings a novel conceptualisation of everyday work practices of resistance against AI through the qualitative investigation of 30 artists and designers in Greater Melbourne, Australia, between 2022 and 2024. Building on resistance theory, I propose to re-position the AI-labour debate towards the forms of everyday resistance creative workers employ to <i>avoid</i>, <i>break</i>, and <i>construct</i> discourses against AI. More specifically, I demonstrate how <i>valuing imperfections</i> and <i>reviving the analogue</i>, amongst other practices, are transformed into acts of resistance against the extractive modes of production and the ideology of frictionlessness that AI catalyses in the creative industries. This article shows that uncovering, explaining, and conceptualising resistance is an important step to further unpack the political economy of Artificial Intelligence.</p></div>","PeriodicalId":47165,"journal":{"name":"AI & Society","volume":"41 9","pages":"8445 - 8460"},"PeriodicalIF":6.1,"publicationDate":"2026-06-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s00146-026-03135-1.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149098307","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Do chatbots change how we think about sexuality? towards an empirically informed philosophy of sexuality in the age of AI-mediated intimacy","authors":"Laurent Voet,&nbsp;Tom Claes,&nbsp;Lode Lauwaert","doi":"10.1007/s00146-026-03132-4","DOIUrl":"10.1007/s00146-026-03132-4","url":null,"abstract":"<div><p>Millions of users now engage in sexual and romantic interactions with AI chatbots, from dedicated companion platforms such as Replika, Ourdream.ai and Character.AI to general-purpose large language models such as ChatGPT, whose developers have recently discussed permitting erotic content for verified adults. These developments constitute what McArthur (2022) terms second-wave sextech, in which the artefact itself becomes the focal point of sexual engagement. Contemporary philosophical debate on second-wave sexual technologies is organized around three dominant positions: instrumentalist, abolitionist feminist, and polysemic. Each is primarily concerned with the ethical permissibility of their design and use. We argue that all three share a significant shortcoming: they fail to address the conceptual pressure that AI-mediated sexual technologies exert on established categories within the traditional philosophy of sexuality and on everyday concepts people use to make sense of their sexual subjectivity, activities and relations. Current philosophical discussions neglect the question of how these technologies may require us to revise such notions as sexual activity, sexual partnership, intentionality, objectification, and the ontological status of technologically mediated sexual experience. This neglect arises from an insufficient appreciation of the co-constitutive relationship between sexual technologies and the concepts through which we understand sexuality. Drawing on a work in conceptual engineering and techno-moral change, we demonstrate that earlier sexual technologies have already occasioned substantial conceptual and normative shifts, and that the current proliferation of AI-mediated sexual interactions is likely to intensify them. The paper advances two claims: that philosophical inquiry into AI-mediated sexual technologies must engage the broader tradition in the philosophy of sexuality, particularly regarding the ontology of sexual acts and sexual subjectivity; and that an empirically informed approach, drawing on experimental philosophy and established empirical disciplines, is warranted to investigate how these technologies reshape broader conceptualizations surrounding sexuality.</p></div>","PeriodicalId":47165,"journal":{"name":"AI & Society","volume":"41 9","pages":"8433 - 8444"},"PeriodicalIF":6.1,"publicationDate":"2026-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149098036","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}