research ethics

AI ethics principles remain difficult to operationalize in software development due to their high abstraction and lack of engineering pathways. To address this, we propose a participatory design–driven paradigm for building ethics tools, co-developing with industry practitioners to map normative ethical frameworks into interpretable, embeddable, and extensible guidance aligned with real-world development workflows. We formalize this through ethics framework engineering—translating principles into actionable artifacts—and workflow-integrated modeling, validated via a U.S.-based proof-of-concept in autonomous driving. The project delivers plug-and-play, open-source ethics support tools that demonstrably improve ethical decision-making efficiency and cross-functional team alignment in production settings. This work constitutes the first systematic, structured translation of AI ethics principles into engineering contexts, offering both a methodological framework and practical infrastructure to bridge the ethics–engineering gap.

Small- and medium-sized enterprises (SMEs) struggle to implement conventional medical AI ethics frameworks due to resource constraints and fast-paced development environments. Method: This paper proposes the Scalable Agile Framework for Ethics in AI (SAFE-AI), which deeply integrates ethical governance into agile software development. It introduces a novel scenario-based probabilistic analogy mapping mechanism for responsibility quantification, establishes testable metrics for fairness, transparency, and uncertainty management, and incorporates a lightweight ethics review model to support iterative development. Contribution/Results: Unlike static, compliance-centric approaches, SAFE-AI operationalizes and scales ethical practice through test-driven acceptance criteria, full-lifecycle monitoring, and business-aligned design. Empirical evaluation demonstrates its applicability in organizations lacking dedicated ethics teams, significantly enhancing model trustworthiness and stakeholder confidence.

Ethical interventions often struggle to integrate into technical practice due to a lack of practitioner trust. This study operationalizes cognitive trust as a core construct of ethical intervention through 70 moral imagination workshops, identifying its five dimensions. Combining qualitative analysis, moral scenario simulations, and structured deliberation, the research systematically delineates 23 failure modes. Building on these insights, the project proposes nine design principles that address a critical theoretical gap in existing co-integration frameworks. These principles offer testable failure hypotheses and actionable design guidance, significantly enhancing the relevance and practical efficacy of ethical interventions in real-world technological contexts.

This paper addresses a fundamental ethical dilemma in software development: prevailing approaches emphasize individual-level tools while neglecting structural barriers rooted in organizational practices, socio-technical systems, and governance mechanisms. To redress this imbalance, the study advances an “infrastructure-centered” conceptual framework—replacing the dominant “tool-centered” paradigm—drawing on Science and Technology Studies (STS) and media infrastructure theory to systematically analyze the institutional conditions enabling ethical action. Through interdisciplinary conceptual analysis, it exposes how values become embedded—and constrained—by tacit architectural assumptions and identifies critical intervention points within technical and organizational infrastructures. The work contributes novel theoretical insights and actionable design strategies to HCI, shifting emphasis from individual accountability toward systemic empowerment. Ultimately, it supports the development of sustainable, institutionally embedded ethical practices in technology production. (149 words)

How can users dynamically calibrate trust in automated systems—appropriately relying when the system is correct and promptly rejecting it when erroneous? This study proposes six interdisciplinary design principles, pioneering the systematic integration of pragmatics’ “common ground” theory and Grice’s cooperative principles into human–computer interaction (HCI) design, thereby establishing a dynamic, context-aware framework for credibility perception alignment. Methodologically, it synthesizes cognitive psychology, user experience (UX) design, and ethics, with emphasis on transparency and communicative effectiveness. Contributions include: (1) the first translation of foundational pragmatic theories into actionable, HCI-oriented design heuristics; (2) a structured, empirically grounded guideline supporting precise trust assessment; and (3) demonstrable improvements in human–AI collaboration safety, efficiency, and user satisfaction, alongside applicability to diagnostic evaluation of existing systems’ trustworthiness.

Social challenge studies—such as online experiments exposing participants to harmful content—lack systematic ethical guidelines; risk mitigation and oversight mechanisms remain markedly underdeveloped compared to established frameworks in medical challenge research. Method: This paper systematically adapts the mature ethical framework of medical challenge studies to computational social science, integrating interdisciplinary analysis to develop context-sensitive ethical principles for online environments and proposing a novel assessment mechanism for long-term latent harms. Contribution/Results: It establishes the first operational ethical standards system specifically designed for social challenge research, thereby addressing a critical regulatory gap. It advances institutionalized ethics review processes tailored to digital experimentation and catalyzes scholarly discourse on risk governance in digital research contexts. By bridging disciplinary divides, the work provides actionable guidance for researchers, ethics boards, and platform partners navigating ethically complex online interventions.

This study addresses the significant gap between prevailing ethical principles for trustworthy artificial intelligence and their practical implementation, as existing frameworks often remain abstract and lack operational guidance. Leveraging the OECD dataset, this work presents the first systematic mapping and comparative analysis of global trustworthy AI tools and certification mechanisms, empirically examining dimensions such as ethical coverage, lifecycle integration, stakeholder engagement, and tool typology. The findings reveal an overemphasis on fairness, transparency, and robustness, while critical aspects like explainability, digital security, and environmental sustainability are largely neglected. Moreover, current tools predominantly target late-stage development phases, offering insufficient support for early design processes and educational policy. To bridge these gaps, the study proposes a governance pathway that broadens ethical objectives, spans the entire AI lifecycle, and strengthens multi-stakeholder collaboration, thereby offering structural insights for institutionalizing trustworthy AI.

This study addresses the limitations of current antisocial behavior interventions, which predominantly rely on punitive measures and lack integration of technology and ethical considerations, thereby failing to foster public responsibility and preventive awareness. Reframing the issue as a human-computer interaction challenge, this work proposes a lightweight digital intervention system that embeds an ethics-informed framework—derived from public opinion in the UK—into its design. The system features a QR code–based reporting interface and an online awareness course. Validation through structured interviews and online surveys demonstrates that the approach effectively enhances public participation, education, and shared responsibility without displacing existing punitive mechanisms. By balancing deterrence with prevention, the proposed solution offers governments a scalable, technology-enabled complement to conventional governance strategies.

Cybersecurity researchers often lack actionable frameworks for stakeholder-oriented ethical analysis. This paper proposes a systematic stakeholder analysis methodology that categorizes stakeholders into four archetypal groups—primary users, secondary affected parties, governance entities, and the general public—and maps them onto empirical research techniques including semi-structured interviews, scenario modeling, and risk mapping, illustrated with real-world case studies. The framework bridges a critical gap in cybersecurity ethics practice by enabling methodologically grounded, context-sensitive ethical assessment. Evaluation demonstrates that research teams using the framework achieve significantly improved efficiency and accuracy in identifying ethical risk exposure points across the project lifecycle, thereby enhancing the rigor, reproducibility, and practical applicability of ethical review. It notably strengthens systematicity in ethical reasoning, improves risk detection precision, and supports more robust, evidence-informed ethical decision-making.

This study addresses the frequent neglect of environmental impacts in computationally intensive research—such as artificial intelligence—due to ambiguous ethical review policies. It presents the first systematic framework integrating environmental sustainability into the ethical oversight of computational research. By delineating clear review boundaries, establishing evidentiary standards, and developing researcher self-assessment tools, the framework enables institutional ethics committees to effectively evaluate the environmental costs of proposed projects. This approach provides actionable guidance for ethical review processes and encourages researchers to proactively consider the ecological footprint of their work during early design stages, thereby addressing a critical gap in current research ethics frameworks concerning sustainability.