Abstract
Abstract Misinformation poses one of the most urgent challenges of our society and raises the question of how to deal with it and manage its rapid spread. To address this problem, a promising approach relies on AI-based misinformation detection. This chapter of the book offers a critical analysis of the ethical implications associated with the design, deployment, and use of misinformation detectors (MDs). Designing and deploying an MD—an AI system that automatically identifies misinformation—is a complex undertaking that requires an interdisciplinary approach, as the challenges faced by MD designers and deployers encompass not only technical aspects, but also linguistic, sociological, political, and especially ethical dimensions. Our analysis is ethics-oriented and follows two main lines of inquiry: (1) Ethics by Design, which focuses on issues related to the design process of an MD, and (2) Ethics of Impact, which addresses the intended and unintended effects of MD deployment and use.

Abstract
As Artificial Intelligence (AI) systems increasingly permeate sensitive domains such as finance, healthcare, and media, ensuring their ethical deployment has become a central concern for researchers, policymakers, and practitioners. Current auditing tools often assess isolated principles, such as fairness or explainability, lacking a comprehensive view of the ethical risks involved. This paper presents a multidimensional framework for ethical evaluation of AI systems, designed to support responsible AI governance and alignment with the United Nations Sustainable Development Goals (SDGs). The proposed approach enables the simultaneous analysis of key ethical dimensions, including fairness, bias, explainability, robustness, transparency, and legal compliance. We demonstrate the applicability of this tool through one extensive case study: a credit scoring system, considered high-risk under the AI Act. This work contributes to operationalizing responsible AI governance, providing insight for policymakers, regulators, and practitioners to ensure ethical, legally compliant, and socially responsible AI deployment.

Abstract
Education 5.0 is a new paradigm in education posing many challenges and opportunities. This paper uses qualitative methods to explore students' and teachers' experiences with online learning to understand the challenges, benefits, and vision for a successful blended learning model, proposing a dynamic framework for blended learning. Results of in-depth interviews show the three main challenges of blended learning: pedagogical design, technological design, and environment/ setup design. Finally, the study discusses insights into future directions for developing Education 5.0, including the need for ongoing research, collaboration communities, curricula personalization, and innovation in the field.

Abstract
This study examines the scientific contributions of the Higher Institute of Accounting and Administration at the University of Aveiro (ISCA-UA) from 2019 to 2022, focusing on how these align with Education 5.0 and Society 5.0 goals. Using a case study approach, data were collected from institutional records, analyzing publications by type and thematic focus, emphasizing areas that promote societal well-being, multiliteracy, and educational innovation. The methodology involves a mixed-methods approach: quantitative analysis assesses publication trends, distribution by faculty rank, and output frequency, while qualitative analysis identifies themes relevant to societal and educational advancements. This approach provides insights into how ISCA-UA's research aligns with Education 5.0 objectives, fostering both technical and socio-emotional skills needed for a super-smart society. Findings highlight an increase in publications addressing digital transformation, sustainability, and governance, reflecting the institution's adaptability and responsiveness to societal shifts, particularly noticeable during the COVID-19 pandemic. This emphasis supports Education 5.0s aims of preparing students with versatile skills for modern challenges. The study contributes to the academic literature by showing how higher education institutions can align research outputs with global educational frameworks, promoting interdisciplinary skills and social responsibility. Future research could explore the impact of these themes on curriculum design and student development, further supporting the evolution toward Education 5.0.

Abstract
Manual selection of real estate properties can pose considerable challenges for agents since it needs a careful balance of various factors to satisfy client requirements while also manoeuvring through the complexities of the market. Although automated valuation models are widely used to estimate property market values, they are not designed to support property recommendation tasks. To address this gap, filteringbased recommendation methods have been explored, including collaborative and content-based approaches. However, these methods face several limitations in the real estate domain. This paper proposes a recommendation methodology designed to identify houses that closely resemble a given property, allowing agents to select the best matches based on geographical and physical characteristics. To assess the performance of the proposed methodology, we employ a range of evaluation metrics that measure different aspects of the model's effectiveness in ranking and recommending relevant items. The findings suggest that, while geographic features may slightly influence ranking behaviour, the model is capable of producing diverse and relevant recommendations consistently.

Abstract
Data leakage is a critical issue in deep learning inflating performance and compromise validity, especially in sensitive areas like medical imaging. This study systematically evaluates two common leakage types in oral squamous cell carcinoma classification from histopathology images: (1) preprocessing leakage (global normalization before dataset splitting) and (2) a severe sample-related (patient-related) contamination scenario created by mixing closely related original and augmented images across splits. We trained 11 CNN and Transformer-based models on a public oral cancer histopathology dataset, benchmarking results against published leakage-free baselines. The results obtained show that the configuration with random splitting of original and augmented images (Scenario 2) artificially increased accuracy by up to 18% (mean +14.3%) compared to leakage-free conditions, while the preprocessing-based leakage (Scenario 1) showed smaller deviations (+1.8%). These inflated metrics arise from a combination of cross-split contamination between closely related samples and increased dataset redundancy, rather than genuine gains in generalization ability. Transformers improved leak-free accuracy (+3.9%) but degraded performance in Scenario 2 (-1.4%), revealing sensitivity to sample-specific biases. The observed performance gains under data leakage conditions are methodological artifacts that undermine clinical reliability, with a severe sample-related contamination scenario (Scenario 2) with random splitting of original and augmented images being particularly detrimental due to its promotion of non-generalizable feature learning. The quantitative benchmarks established here-including a mean accuracy gap of 12.5% (Scenario 2 vs. Scenario 1) across 11 models and Transformer architectures’ sensitivity to contamination-reveal fundamental tradeoffs between metric inflation and model trustworthiness. These findings establish quantitative benchmarks for leakage impacts in medical imaging and inform future guidelines for trustworthy AI development in pathology. © The Author(s) 2026.