Beyond Procedural Compliance: Human Oversight as a Dimension of Well-being Efficacy in AI Governance

Beyond Procedural Compliance: Human Oversight as a Dimension of Well-being Efficacy in AI Governance Yao Xie School of Medicine, University College Dublin Dublin, Ireland toyaoxie@gmail.com Walter Cullen School of Medicine, University College Dublin Dublin, Ireland walter.cullen@ucd.ie Abstract Major AI ethics guidelines and laws, including the EU AI Act, call for effective human oversight, but do not define it as a distinct and developable capacity. This paper introduces human oversight as a well-being capacity, situated within the emerging Well-being Efficacy framework. The concept integrates AI literacy, ethical discernment, and awareness of human needs, acknowledging that some needs may be conflicting or harmful. Because people inevitably project desires, fears, and interests into AI systems, oversight requires the competence to examine and, when necessary, restrain problematic demands. The authors argue that the sustainable and cost-effective development of this ca- pacity depends on its integration into education at every level, from professional training to lifelong learning. The frame of human oversight as a well-being capacity provides a practical path from high-level regulatory goals to the continuous culti- vation of human agency and responsibility essential for safe and ethical AI. The paper establishes a theoretical foundation for future research on the pedagogical implementation and empirical validation of well-being effectiveness in multiple contexts. 1 Introduction The contemporary world is entering an era defined by artificial intelligence (AI) (Xu et al., 2024). Driven by rapid innovation, AI transforms how people learn, work, communicate, and decide (Abuzaid, 2024; Afroogh et al., 2024). This movement marks an irreversible shift toward AI-mediated environments where intelligent systems increasingly shape everyday life and well-being(Singh & Tholia, 2024). The transformation brings new opportunities for creativity and efficiency but also exposes people to losses in agency, coherence, and collective trust. The speed of technological change, the weakening of human voice, and public indifference toward data protection reveal how easily individuals trade agency for convenience (Yatani et al., 2024). These conditions make human oversight not only a technical necessity but also a fundamental capacity for collective well-being (Corrêa et al., 2025; Langer et al., 2024; Sterz et al.). Human oversight has become a central principle of global AI governance (Koulu, 2020). The European Union AI Act, the OECD AI Principles, and the UNESCO Recommendation on the Ethics of Artificial Intelligence all emphasise its role in keeping technology aligned with human priorities (Enqvist, 2023). These frameworks recognise that AI should not operate without human judgment. Second Conference of the International Association for Safe and Ethical Artificial Intelligence (IASEAI’26). However, their practical interpretations often focus on institutional examples: a clinician approving an algorithmic diagnosis, a financial analyst authorising an automated transaction, or a teacher monitoring AI-assisted assessment (Koulu, 2020). These examples describe regulated professional contexts but fail to reflect the broader scope of AI influence such as the subtle ways in which design and daily use shape human experience (Koulu, 2020). Most people encounter AI in ordinary contexts rather than in professional ones. They interact with recommendation systems that shape preferences, social media feeds that influence belief, and digital platforms that collect and reuse personal data in what can be described as hybrid space— ‘refer to merging physical and digital spaces’(de Souza e Silva et al., 2025). Public institutions also increasingly depend on algorithmic tools in police, welfare allocation, and recruitment. In such environments, oversight becomes diffuse and largely invisible (Koulu, 2020; Sterz et al.). Current regulations seem to assume that people already have the awareness and critical discernment needed to make informed choices. In reality, many people engage with AI outputs passively, lacking the reflective capacity necessary for effective oversight (Chen et al., 2023). Humans also have inherent limitations in decision making, often relying on cognitive shortcuts that reduce awareness and deliberation (Curran, 2015; Dale, 2015; Yoder & Decety, 2018). The governance challenge extends beyond technical regulation to the cultivation of human awareness and agency across all levels of society (Sigfrids et al., 2023; Van Popering-Verkerk et al., 2022). Governance cannot depend solely on external rules or audits, it requires stronger human capacity (Enqvist, 2023; Yeung et al., 2020). As AI becomes embedded in daily life, the question is not how a single person supervises a specific system, but how human agency can be maintained in complex, distributed, and rapidly evolving environments. This challenge calls for scalable human oversight—the capa

This content is AI-processed based on open access ArXiv data.