Search Results (132)

Background/Objectives: Lung cancer is a leading cause of cancer-related mortality worldwide. As patients increasingly utilize large language models (LLMs) for health information, evaluating the readability and patient-centeredness of these tools is critical. This study aims to compare the performance of ChatGPT-4o mini, Microsoft Copilot, and Google Gemini in providing lung cancer information, focusing on their utility for individuals with limited health literacy. Methods: In this cross-sectional study (March 2026), 30 responses to ten standardized lung cancer-related queries were analyzed. Outputs were assessed using JAMA benchmarks and mDISCERN for quality, the SMOG index for readability, and PEMAT-P for understandability and actionability. Inter-rater reliability was analyzed using intraclass correlation coefficients (ICCs). Results: ChatGPT-4o mini demonstrated superior readability, achieving a sixth-grade level (SMOG: 6.23 ± 0.72, p < 0.001). Gemini achieved higher JAMA scores, indicating stronger academic rigour. While PEMAT-P scores were highest for ChatGPT (63.7%), all models exhibited moderate mDISCERN quality. Inter-rater reliability was excellent for JAMA (ICC = 1.000) and PEMAT-P (ICC = 0.883), though moderate for mDISCERN (ICC = 0.365), reflecting inherent interpretative subjectivity in qualitative assessment. No hallucinations were observed. Conclusions: Current LLMs exhibit a trade-off between accessibility and academic rigour: ChatGPT favours patient-friendly readability, while Gemini emphasizes structured content. The observed inter-rater variability in mDISCERN underscores the complexity of standardizing qualitative AI evaluation. These findings suggest that LLMs function best as complementary aids rather than substitutes for physician-led communication. Full article

This study examines the phenomenon of pseudoconfident knowledge in the context of the everyday use of generative artificial intelligence. By pseudoconfident knowledge, we mean a response that is substantively plausible, rhetorically coherent, and outwardly persuasive but is treated and understood as knowledge before its actual reliability has been established. Of course, we do not use the term “pseudoconfident knowledge” to denote knowledge in the strict epistemological sense. Rather, it denotes a special form of AI-generated content that acquires the status of knowledge in the user’s perception before its reliability, source-based justification, or factual correctness have been established. The problem here is not that such an answer is already knowledge but that it is prematurely accepted as knowledge because of its coherence, completeness, and rhetorical confidence. The aim of the study is to identify the epistemic gap between the everyday operational integration of artificial intelligence and the user’s critical ability to distinguish between persuasiveness and justification. The theoretical framework combines approaches to AI literacy, epistemic vigilance, and contemporary forms of digital mediation in the circulation of knowledge. The empirical basis of the study is an online survey of AI users. The analysis was conducted using descriptive statistics, contingency tables, and methods for testing associations between categorical variables. The results show that the key differentiating factor is not the frequency of AI use, but the strategy used in handling its responses. More epistemically robust positions are associated with practices of comparison, editing, and verification, whereas uncritical acceptance of the answer is associated with greater vulnerability to pseudoconfident knowledge. We conclude that the spread of generative artificial intelligence is producing a new socioepistemic problem that calls for a shift in emphasis from simple instrumental literacy toward a culture of verification, doubt, and epistemic responsibility. Full article

This study presents the development and psychometric validation of the Artificial Intelligence Competence and Literacy Test (CAIA), designed to assess artificial intelligence (AI) literacy and competence in higher education contexts. As AI becomes increasingly integrated into academic and professional environments, reliable instruments are needed to evaluate individuals’ conceptual understanding, ethical awareness, and applied competencies related to AI. The instrument was administered to a sample of 510 university students from several faculties at a Spanish university. Exploratory and confirmatory factor analyses were conducted using a cross-validation design. Results supported a multidimensional structure, in the final 18-item version of the instrument, composed of two correlated factors (critical—conceptual AI literacy and creative—applied AI competence) and a second-order hierarchical model representing a global CAIA score. Model fit indices were acceptable-to-good, and reliability estimates, including ordinal coefficients and measurement error indicators, showed adequate precision for both individual and group-level interpretation. Evidence of construct validity was further supported through convergent and discriminant analyses, as well as hypothesis testing across academic subgroups. The findings suggest that the CAIA provides a theoretically grounded and psychometrically robust instrument for assessing AI-related competencies in higher education. The instrument may support research, curriculum design, and evaluation of educational initiatives aimed at promoting informed, critical, and responsible engagement with artificial intelligence in digitally mediated learning environments. Full article

As intelligent systems become decision-support tools in the architecture, engineering, and construction (AEC) industry, establishing human–AI trust is critical. However, in engineering consulting, the psychological mechanisms underlying trust formation remain unclear. Grounded in Self-Determination Theory and the Stereotype Content Model, this study utilized multi-wave survey data from Chinese engineering consulting employees to investigate these mechanisms. We examined how perceived autonomy influences human–AI trust through the competitive dual-mediation of warmth perception and competence perception, alongside the asymmetric moderating role of critical thinking. Results reveal that perceived autonomy directly enhances trust. However, social cognition acts as a competitive mechanism: autonomy positively impacts trust via warmth perception but generates a negative indirect effect via competence perception. Furthermore, critical thinking exerts an asymmetric boundary effect; it does not interfere with the intuitive warmth pathway but significantly intensifies the negative indirect effect through the competence pathway. Ultimately, these findings highlight that perceived autonomy exerts a double-edged sword effect in the context of human–AI collaboration. To mitigate professional defensive rejection and calibrate trust, AEC firms should prioritize human-in-the-loop deployment strategies, objective interface designs, and the cultivation of AI collaborative literacy. Full article

Guided by Social Cognitive Theory (SCT) and the Technology Acceptance Model (TAM), this qualitative study examines how K–12 teachers across five countries, the United States (n = 7), India (n = 5), Qatar (n = 5), Colombia (n = 5), and the Philippines (n = 4), conceptualize AI literacy and integrate generative AI into their practice. Through 26 semi-structured interviews conducted in summer and fall 2025, we identified three cross-national patterns that challenge dominant narratives about AI adoption in education. First, institutional support did not uniformly predict AI literacy depth: the four Filipino teachers developed sophisticated prompt engineering competencies despite low institutional backing, while the five Indian teachers showed the lowest awareness despite strong organizational support. Second, prompt engineering awareness functioned as a critical differentiator between teachers who engaged with AI as a pedagogical skill and those who treated it as an opaque productivity tool. Third, AI use for lesson preparation far outpaced classroom-facing application across all contexts. These findings reframe AI readiness as a question not of access and support but of whether conditions cultivate the interaction competence that meaningful integration demands. Full article

Understanding student artificial intelligence (AI) literacy in the context of higher education is crucial as technology advances and AI use increases. The purpose of this study is to better understand how university students perceive, define, and apply AI literacy within their own educational experiences and from their own disciplinary lens. Collecting electronic survey responses from 130 graduate and undergraduate students across several disciplines including First-Year Writing, Communication Studies, and Education, this study attempts to elucidate how students articulate and perceive their own degree of AI literacy—Access, Understanding, Critical Thinking, Application, and Ethics—in the educational context. Overall, students reported infrequent use, using ChatGPT most often. Education students reported a lower understanding of AI than non-education students. Undergraduates reported higher rates within ethics than graduate students. No significant differences in AI literacy were found between students who were or were not first-generation students, students who did or did not receive financial aid, or by gender. Students reporting higher rates of use also reported higher rates of AI literacy. Crucially, this study provides key qualitative and quantitative insights exploring how students perceive their own AI literacy. Understanding the current state of students’ AI literacy is important to facilitating holistic student success in academic environments and career readiness as institutions of higher education adapt and prepare curricula, programs, and interventions addressing AI literacy across disciplines. Full article

As artificial intelligence reshapes professional workflows, understanding what drives effective AI use among employees has become a critical concern for organizations. Moving beyond traditional technology acceptance frameworks, this study develops an integrative multi-level model to examine the behavioral determinants of AI use performance (AUP) among journalists. Drawing on the Technology Acceptance Model (TAM) and the Expectation Confirmation Model (ECM) and incorporating individual and organizational factors, a survey was conducted among 543 journalists in China. Hypotheses are tested via a hybrid PLS-SEM and artificial neural network (ANN) approach to capture both linear and non-linear relationships. The findings reveal that expectation confirmation significantly enhances AUP by driving perceived usefulness and satisfaction. Digital literacy, personal trust in AI, and organizational support positively influence AUP, whereas communication barriers exert the strongest negative effect. Demographic variables (gender, age, education) show no significant impact. Notably, the ANN sensitivity analysis identifies communication barriers as the most influential predictor overall, a finding not apparent from linear analysis alone. This study advances theoretical understanding of employee behavioral responses in AI-integrated professional contexts and offers practical insights into how organizations can foster effective employee–AI collaboration through targeted communication strategies and supportive infrastructure. Full article

The rapid expansion of generative artificial intelligence (AI) is transforming higher education and creating new opportunities that are associated with the development of perceived professional competencies. At the same time, the accounting profession increasingly requires graduates who can evaluate sustainability disclosures and form informed judgments regarding sustainability-related information. However, limited research has examined how AI-supported learning relates to sustainability-oriented decision-making capabilities in accounting education. Drawing on Decision Support Systems (DSS) theory and constructivist learning theory, this study examines the associations between generative AI-supported learning and students’ perceived sustainability judgment capability. Specifically, the study investigates the mediating roles of perceived critical thinking and perceived sustainability knowledge, as well as the moderating role of AI literacy. A quantitative, cross-sectional research design was employed using self-reported survey data collected from 721 accounting students, and the proposed relationships were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The findings indicate that generative AI-supported learning is positively associated with students’ perceived critical thinking and perceived sustainability knowledge. In turn, both constructs show significant positive relationships with perceived sustainability judgment capability, with perceived sustainability knowledge demonstrating a stronger association. Additionally, AI literacy strengthens the relationships between generative AI-supported learning and the cognitive constructs. Importantly, the study captures students’ self-reported perceptions of their cognitive and judgment-related capabilities and does not assess objective cognitive performance or demonstrated judgment ability. The study contributes to the literature by positioning generative AI as an educational decision-support mechanism associated with perceived sustainability-oriented judgment capability through cognitive pathways, while highlighting the importance of aligning theoretical claims with perceptual measurement approaches. Full article

This qualitative review of the literature explores current conversations about the impact of Artificial Intelligence (AI) on teacher education in general and pre-service teachers in particular. Recent advances in AI are beginning to influence teacher education, where curricula, practicum, and school field experience now incorporate AI in curriculum-based instruction and as a context for teaching digital literacy, not as an isolated tool. Researchers regularly situate these shifts alongside broader educational practices and policy. There is also substantial literature dealing with pressing ethical and practical questions such as data privacy and algorithmic bias, equitable access to technology, and the challenges experienced by under-resourced schools. Together, these studies indicate that teachers are redefining and reconfiguring both their own teaching and teacher education, enabled by AI in new, more flexible and responsive ways. Within this shifting paradigm, pre-service and in-service teachers are not conceived as mere end-users but as reflective practitioners who take up such tools, critically question their ramifications, and, sometimes, lead the way in utilizing AI in educational practice, including mainly pedagogical practices. To explain the shared components identified in the present review, this paper offers a post hoc conceptual synthesis of eight recurring dimensions of sustainable AI integration in teacher education. Full article

Background: Effective communication with patients is vital for improving health outcomes in chronic disease management. In this study, we investigated WoundScribeAI’s Scribe AI, also known as Ambient Technology, and its patient education and engagement app, Pingoo.AI. It employed a multi-agent AI model that leveraged Large Language Models (LLMs) and NotebookLM to enhance patient communication in clinical settings. Methods: The system comprised specialized agents that transcribed healthcare provider–patient conversations through ambient dictation. This transcription generated medical notes that followed the Subjective, Objective, Assessment, and Plan (SOAP) format—a structured document used by healthcare providers to record and communicate information about patient encounters. Simultaneously, comprehensive visit summaries were also created. In the next step, these visit summaries were used to produce conversational and educational content by leveraging NotebookLM, an AI model introduced by Google that can generate podcast-style conversations from provided information. Integrating these agents allows clinicians to deliver engaging, empathetic, and actionable information to patients. Medical experts conducted a two-phase evaluation of the system’s performance based on multiple criteria, with a particular focus on diabetes education and diabetic foot care. The first phase used pre-recorded training videos, while the second phase involved simulated consultations by clinicians using the system. To validate the AI-generated educational content, we used several established frameworks in health communication that closely align with our enhancement goals. Results: The results showed that the AI model generated accurate clinical documentation and met the criteria for accurate SOAP Notes, visit summaries, and engaging educational content for patients. Given that hallucination is a significant concern related to large language models, especially in critical fields like healthcare, we meticulously analyzed the generated outputs to identify any signs of hallucinated information. Three outcomes successfully passed the validation criteria, including accuracy, completeness, comprehensiveness, absence of potential harm, and no hallucination. Additionally, the Conversational Education content was confirmed against established patient education frameworks and met criteria such as the use of metaphors, empathetic tone, and appropriate language, providing additional detail to help manage the condition. Conclusions: By providing specific instructions and prompts to NotebookLM to transform visit summaries into educational conversations, we significantly enhanced the comprehensiveness and engagement of the content for patients. In contrast to a traditional summary of the clinical visit, the podcast-style conversation enriched the content with background information, encouraging language, an empathetic tone, and helpful metaphors. Our analysis confirmed that the system did not exhibit any hallucinations, highlighting the effectiveness of our approach in mitigating this risk. These findings support the use of multi-agent AI models, combined with ambient dictation and tools like NotebookLM, to improve patient communication that surpasses traditional paper-based brochures, which are often impersonal, minimal, and do not always adhere to recommended factors for health literacy. Full article

Artificial intelligence (AI) is reshaping higher education by changing how students access knowledge, complete academic tasks and engage in innovation. At the same time, unequal access to AI-related competencies may reproduce existing educational inequalities, which raises important concerns for Sustainable Development Goal 4 (SDG 4). Drawing on cultural capital theory and research on digital inequality, this study examines whether family cultural capital is associated with university students’ innovative capacity through AI literacy. In this study, AI literacy is defined as students’ ability to understand, evaluate and use AI critically and responsibly across different contexts. Survey data were collected from 1020 Chinese university students and analyzed using structural equation modeling (SEM) with split-sample validation. The results indicated that family cultural capital remained significantly associated with innovative capacity although its two dimensions operated differently. Cultural resources had a significant direct effect on innovative capacity and also positively predicted technical application skills but not awareness of the social impact of AI. Embodied cultural capital did not have a significant direct effect on innovative capacity, but its total effect was significant, and it positively predicted both dimensions of AI literacy. Mediation analysis further showed that technical application skills significantly mediated the relationship between both dimensions of family cultural capital and innovative capacity, whereas awareness of the social impact of AI did not show a significant mediating effect. These findings suggest that family cultural capital continues to matter in the AI era not only through direct advantage but also through its conversion into AI-related competencies. The study highlights the need for higher education institutions to strengthen equitable support for practical AI capability development in order to promote inclusive innovation and advance SDG 4. Full article

The expanding role of artificial intelligence (AI) in education raises important questions about how AI-supported learning may foster higher-order thinking and creative talent development. Guided by social cognitive theory, the current research examined how AI self-efficacy predicts creativity among second language (L2) learners through the mediating roles of AI literacy and critical thinking disposition. Two substudies were conducted. Study 1 (N = 72) tested a simple mediation model and demonstrated that AI self-efficacy positively predicted creativity both directly and indirectly through AI literacy. Study 2 (N = 135) extended these findings by incorporating critical thinking disposition and by using another measure of creativity. Results showed that AI self-efficacy positively predicted creativity, and this relationship was mediated independently by AI literacy and critical thinking disposition, as well as sequentially through both factors. The current study provides empirical evidence for pathways linking AI self-efficacy, AI literacy, critical thinking disposition, and creativity in AI-supported L2 learning. It highlights the importance of reflective and critical use of AI tools in language education. Full article

Generative AI (GenAI) is reshaping second language (L2) writing not only by altering how learners generate, revise, and refine text but also by changing how writers justify, disclose, and remain accountable for AI-mediated decisions. Yet much prompt-literacy work still treats prompting as output optimization or leaves it under-theorized as a general ability to “use AI well.” This conceptual article addresses that gap by reconceptualizing Fine-Tuned Prompt Literacy (FTPL) as an interaction-first learning-and-accountability framework for GenAI-mediated L2 writing. We argue that prompt literacy should be understood not simply as better prompting, but as the trained ability to set communicative and genre constraints, interrogate provisional AI outputs, corroborate claims, revise prompts and texts iteratively, and document accountable uptake decisions. To clarify FTPL’s theoretical distinctiveness, we position it in relation to AI literacy, critical GenAI literacy, and prompt literacy research, and define four interlocking dimensions—learner empowerment, prompt optimization, critical evaluation, and ethical responsibility. We further operationalize the framework through observable interactional indicators, process evidence, and assessment/accountability implications relevant to instructional and institutional contexts. By reframing prompt literacy as a genre-sensitive and ethically accountable interactional competence, this article offers a conceptual model for studying and designing GenAI-mediated writing beyond product improvement alone. Full article

Africa has the world’s youngest population, and many young adults rely on informal or temporary employment, making digital financial literacy (DFL) critical for long-term financial resilience and sustainable economic development. In this paper, we present findings from a two-phase mixed-methods study. In Phase 1, we surveyed 300 Rwandans aged 18–32 on financial knowledge, digital skills, and financial behaviors to explore key gaps in DFL. Results show modest financial knowledge and moderate digital literacy, with common budgeting and saving practices but key cybersecurity awareness-practice gaps. Gender and education disparities are also evident. To address the low loan literacy observed in Phase 1 findings, we conceived an AI-enabled mobile money loan literacy chatbot and explored user interactions with the chatbot, along with perceived usability and usefulness in Phase 2. Our findings highlight design considerations for promoting intention to adopt DFL interventions. The study aligns with the United Nations Sustainable Development Goals (SDGs) 1 (No Poverty), 5 (Gender Equality), 8 (Decent Work and Economic Growth), 9 (Industry, Innovation and Infrastructure), and 10 (Reduced Inequalities). Full article

Background/Objectives: Echocardiography reports are essential diagnostic tools, but their complexity and specialized English terminology frequently hinder comprehension for non-specialists and patients. This study addresses these accessibility gaps by developing a resource-efficient large language model (LLM) system designed to translate and summarize English echocardiography results into Traditional Chinese. Methods: To overcome significant hardware constraints, we utilized Quantized Low-Rank Adapter (QLoRA) techniques and the Unsloth acceleration framework to fine-tune LLaMA-3.2-1B and LLaMA-3.2-3B-Instruct models on a single mid-tier GPU. The system employs a dual-stage inference architecture: the first stage provides technical medical translation for clinicians, while the second stage generates simplified, patient-centric educational summaries to enhance health literacy. Results: Evaluation across multiple metrics, including BLEU, ROUGE, METEOR, and Perplexity, demonstrated that the LLaMA-3.2-3B-Instruct model with the AdamW 8-bit optimizer achieved the most stable validation performance, excelling in semantic coherence and structural consistency. A preliminary qualitative error analysis conducted in the Discussion section further identified clinical nuances, such as terminology simplification and minor hallucinations, underscoring the critical necessity of a Human-in-the-Loop verification procedure. Conclusions: These findings validate the feasibility of deploying cutting-edge medical AI in resource-limited clinical environments. While the results reflect validation-only performance on a specialized dataset, the platform offers a scalable foundation for enhancing clinical decision support and health literacy through accessible, automated medical text processing. Full article