generative ai

When AI Joins the Diagnosis: How Doctor–AI Collaboration Shapes Perceived Doctor Responsibility Under Perceived Diagnostic Errors

Artificial intelligence (AI), especially generative AI, is becoming deeply embedded in healthcare, making doctor–AI collaborative diagnosis a common mode of medical decision-making. This development raises a critical question about responsibility attribution: When people perceive that a diagnostic error has occurred, how does AI involvement shape patients’ and observers’ judgments of the doctor’s responsibility? Drawing on responsibility attribution theory, we examine this question across five studies—one event-related potential (ERP) experiment and four scenario experiments. We find that when a diagnostic error is perceived, doctor–AI collaborative diagnosis (vs. doctor-only diagnosis) reduces perceived doctor responsibility by increasing perceived shared agency. This responsibility-reducing effect is weaker when the doctor rejects correct AI advice than when the doctor accepts incorrect AI advice. Theoretically, our findings show that responsibility attribution in human–AI collaboration involves two stages: agent identification and responsibility allocation. This account extends responsibility attribution theory to human–AI collaboration and identifies perceived shared agency as a key psychological mechanism underlying responsibility judgments in these settings. Practically, the findings can inform technology deployment, responsibility communication, and governance mechanisms in hospitals, AI firms, and regulatory agencies.

Knowing what AI literacy is, but not yet how to teach it: A rapid review of teachers’ professional development needs and effective principles in higher education

From Digital Transformation to Responsible AI Governance: Public-Facing AI Ethics Readiness in the Kurdistan Region

Generative Artificial Intelligence as a Collaborative Designer for Elementary Computer Science Education

Artificial Intelligence (AI) is becoming ubiquitous in our everyday lives. In the United States, AI has increasingly been recognized as an important part of computer science (CS) standards for elementary school students. Prior research work has shown that AI-related content can be taught at upper elementary grade levels. However, classroom-ready, standards-aligned materials for CS in the context of AI remain limited at the elementary level. This paper showcases how educators may use generative AI to design standards-aligned materials for teaching CS and AI in elementary schools. Specifically, this paper presents a design case in which generative AI (Google Gemini) serves as a collaborative designer to co-author four comic-based stories for grades 4–6. The four comic stories, titled “The AI Journey with Milo and Zada,” “Zappy the AI Robot,” “The AI Adventures,” and “Think Like an Engineer,” are mapped to specific AI standards in the Alabama Digital Literacy and Computer Science (DLCS) Course of Study. This work provides a replicable example of using generative AI to co-design standards-aligned instructional materials with responsible use of AI. This work is also useful for instructional designers, elementary school teachers, and policymakers in elementary computer science education.

Guru Logics After Artificial Intelligence

AI-mediated forms of guruship are proliferating across South Asia and beyond, yet they have scarcely begun to receive sustained scholarly attention. This article takes their emergence as an occasion to revisit the conceptual vocabulary through which guruship has been understood, asking what becomes of ‘guru logics’ after artificial intelligence. AI gurus are not merely technological curiosities; nor are they straightforward successors to human gurus. This article treats them as analytic provocations through which longstanding conceptions of guruship can themselves be reconsidered. Guru studies had already been grappling with many of the problems that AI now makes newly conspicuous: imitation and performativity; presence in absence; distributed and non-human personhood; the mediation of charisma; embodiment and its displacement; succession and charismatic continuity. AI therefore strains an existing conceptual repertoire while also showing what it can still do. The article develops the notions of ‘code-born syncretism’ and ‘ambient guruship’, alongside ‘Transhuman Guru Worlding’ and ‘guru cosmoi’, to explore what happens when the guru becomes generative and responsive. ‘Doctrine by interface’ further draws attention to the ways theological possibilities are shaped through the design and affordances of AI systems, while ‘simulation as method’ considers what experimental encounters with such systems might contribute to the ethnographic study of guruship. These concepts are developed through discussions of devotional recombination, embodiment, programmed continuity of the guru after death or exile, and the political economies within which AI gurus are made and encountered. The article takes up histories in which the guru has already exceeded the bounded human body, becoming distributed through multiple material and mediated forms, and examines how generative interaction and recombination rework these inherited guru logics, while letting these forms scale and personalise in novel ways. What emerges is therefore neither unprecedented rupture nor straightforward continuity, but a recursive transformation of existing forms of guruship through computational mediation. The article offers a conceptual framework and analytic repertoire for an emerging field of ethnographic inquiry rather than a completed ethnography of AI guruship. Its larger proposition is that AI gurus matter for more than what they might tell us about religion in an age of artificial intelligence. They also provide an unusual vantage point from which to return to the conceptual problems that have long animated guru studies and to ask what artificial intelligence allows us to see differently about guruship itself.

Generative model to improve the physical simulation of normalized seismic acceleration

Earthquake physics-based simulations have revealed their accuracy limit in high-fidelity broadband strong ground motion scenario prediction. Earlier attempts leveraging AI tried to fill the gap between the low-frequency high-fidelity numerical simulation and the high-frequency accuracy and realism demanded in earthquake engineering. Our research is built on top of the SeismoALICE solution proposed by Gatti and Clouteau (2020), an AI generative approach that renders broadband (0-20 Hz) single-station accelerograms conditioned by low-frequency physics-based simulation outcomes. First, we developed a novel neural architecture based on Encoder, Decoder, and Discriminators that integrates Conformer's advanced attention techniques (Gulati et al., 2020), stabilising the training and producing realistic output for the generation. This novel architecture is trained according to Adversarial Learning Inference with Conditional Entropy (ALICE, Li et al., 2017). Second, our approach employs a similarity evaluation technique called Hyper-Spherical Loss (HSL) in the time domain and an adaptation of the Focal Frequency Loss (FFL) for time series. Our investigation demonstrates that Conformer architecture outperforms previous approaches in super-resolution of earthquake data, such as the STanford EArthquake Dataset (STEAD) Dataset (Mousavi et al., 2019). We finally showcase the enhanced strong motion synthesizer to predict the seismic response at the Cruas Nuclear Power Plant during the 2019 MW 4.9 Le Teil earthquake, providing insightful perspective for future large-scale applications as a downstream generative pipeline.

The Prompt Richness Index: A Proposed Seven-Dimension Framework for Evaluating AI Text-to-Image Generation in Architectural Design Education

Transforming concepts into architectural designs is challenging, as it requires clear ideas and the ability to visualise them. AI-powered text-to-image tools help designers quickly convert concepts into visual representations, enabling faster exploration of design alternatives. This study introduces a proposed ten-step numerical procedure for assessing the richness of textual prompts submitted to text-to-image generative AI tools within an architectural design studio. Twenty-three architecture students were enrolled in the design studio; twenty-two submitted analyzable text prompts as part of a design assignment requiring AI-assisted conceptual visualisation. Each prompt was scored across seven weighted dimensions (subject specificity, style and medium, composition and framing, lighting and atmosphere, colour and palette, quality modifiers, and negative clauses) to produce a composite Prompt Richness Index (R, scale 0–100). Corresponding AI-generated images were independently scored using a parallel Output Richness Index (O, scale 0–100). Pearson’s r between per-student average R and O yielded r = 0.940 (p < 0.001, 95% confidence interval [0.86, 0.98]), confirming a nearly perfect positive linear relationship. Rich-tier prompts were produced by two students and yielded the most architecturally coherent and visually distinctive outputs. Two students produced Sparse-tier prompts (average R < 30) and consistently received undifferentiated, generically rendered outputs. Two further students scored just above the Sparse threshold but showed similarly limited output differentiation. Class-wide deficits were identified in lighting/atmosphere description and negative clause usage. Eight pedagogical recommendations are derived from the findings to guide prompt learning instruction in AI-integrated design studios.

Mental health interventions: an integrated system based on generative artificial intelligence and virtual reality

Abstract Mental health interventions use therapeutic techniques to support emotional well-being, often leveraging emerging technologies to improve accessibility and effectiveness. In recent years, tools such as VR and AI have been increasingly used to create immersive and adaptive therapeutic experiences. This paper presents a framework that integrates generative artificial intelligence (GenAI) and virtual reality (VR) to enhance mental health interventions. The system allows for near real-time co-creation of three-dimensional objects during therapy sessions and provides a customizable “safe space”for emotional regulation. In a user study (n = 30), participants completed anxiety (GAD-7) and depression (PHQ-9) assessments, generated virtual objects linked to specific emotions, and rated the system’s usability (SUS) and recommendability (NPS). The framework achieved a high SUS score of 79 and an NPS of 40, indicating robust usability and willingness to recommend. Notably, neither anxiety nor depression levels, nor prior experience with VR or GenAI, significantly affected these measures. These findings highlight the feasibility and acceptability of combining GenAI and VR to offer immersive, personalized, and user-friendly therapeutic environments, representing a promising direction for future clinical validation in people-centered healthcare programs.

From creation to dissemination: a lifecycle-based evaluation of AI guidance policies in health-related journals

Abstract Background The rapid integration of generative artificial intelligence into health sciences research has prompted major advisory bodies to establish ethical guidelines governing AI use in scholarly publishing. However, practical implementation at the journal level remains inconsistent, with considerable incongruence observed between individual journal policies and overarching publisher mandates. This study conducts a comprehensive analysis of AI guidance policies in international health-related journals to evaluate the current regulatory landscape and provide a robust framework for future guidance. Methods A mixed-methods approach was employed, integrating qualitative subject analysis, inferential statistics, and Multiple Correspondence Analysis (MCA). Through qualitative subject analysis, an attribute-based regulatory schema spanning the research lifecycle was developed. Policy texts were independently evaluated using a dual-coder protocol and transformed into a standardized categorical dataset. Bivariate inferential statistics and MCA were subsequently applied to evaluate structural interrelationships between these regulatory stances and institutional characteristics. Results Descriptive analysis revealed profound regulatory disparities. While a robust consensus exists on prohibiting AI authorship (72.6%), severe regulatory vacuums persist in upstream research phases, with 54.2% of journals omitting guidance on methodological design. Bivariate analyses demonstrated that policy rigor is significantly determined by policy presentation mode (the presence and origin of a journal's policy) and journal prestige. Furthermore, MCA extracted three distinct regulatory typologies: Silent Vacuum prevalent among lower-ranked journals; Restrictive Hardliners enforcing strict bans on visual and methodological AI applications; and Autonomous Pragmatics, where journals with independent policies champion disclosure-based integration. Conclusion This study reveals a polarized AI regulatory landscape across health-related journals, wherein institutions either maintain a complete policy vacuum or implement comprehensive frameworks. Policy presentation mode and journal prestige emerged as the primary structural determinants of regulatory content, with publisher-derived policies and high-impact journals consistently associated with more rigorous guidance. This study advances the theoretical understanding of scientific guidance by conceptualizing AI policy adoption as a complex interplay between administrative standardization and journal prestige. Practically, the proposed lifecycle-based framework serves as a vital diagnostic tool for editorial boards to address critical regulatory gaps, while simultaneously providing authors with a comprehensive guide to navigate the ethical application of generative technologies across the research process.

The Evolution of Artificial Intelligence in Antibody Design: From Structure-Based Engineering to Generative Models

Background/Objectives: Artificial intelligence (AI) has transformed computational antibody engineering by enabling accurate prediction of antibody structures, rational optimization of therapeutic properties, and de novo antibody design. Recent advances in deep learning, protein language models, and generative AI have fundamentally changed the way antibodies are discovered and engineered. This review aims to present the historical evolution of computational antibody engineering, from early structure-based design strategies to modern AI-driven approaches, while highlighting the major computational tools, publicly available databases, current limitations, and future directions of the field. Methods: A comprehensive narrative review of the literature was conducted using PubMed, Scopus, Web of Science, and Google Scholar. Original research articles, methodological studies, and review papers published between 1985 and 2026 were evaluated. Publications were selected according to their scientific relevance, methodological quality, and contribution to the historical development of computational antibody engineering. Results: The review describes the progression of antibody engineering from phage display and structure-based computational methods to machine learning, deep learning, protein language models, and generative artificial intelligence. It summarizes key public databases supporting antibody research, discusses advances in antibody structure prediction and developability assessment, and reviews recent generative models capable of designing antibody sequences and structures. Current challenges, including limited experimental validation, dataset bias, prediction of highly flexible regions, model interpretability, and clinical translation, are also discussed. Conclusions: Artificial intelligence has fundamentally reshaped computational antibody engineering by integrating sequence, structural, and functional information into increasingly accurate predictive and generative frameworks. Although important challenges remain, recent developments indicate that AI-driven approaches will play an increasingly central role in the discovery and optimization of next-generation therapeutic antibodies.

Generative AI as a “peer” in EAL writing feedback

The Brightest Physician in the Room: AI, Professional Recognition, and the Experience of Knowing