Ethical and Inclusive Artificial Intelligence for Sustainable Knowledge Societies: A Global Governance Framework through a Systematic Literature Review

Ethical and Inclusive Artificial Intelligence for Sustainable Knowledge Societies: A Global Governance Framework through a Systematic Literature Review | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Ethical and Inclusive Artificial Intelligence for Sustainable Knowledge Societies: A Global Governance Framework through a Systematic Literature Review Dwi Mariyono This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8230657/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Purpose: To bridge the persistent gap between AI ethics principles and educational practice by proposing a Pedagogically-Centered AI Governance (PCAG) model that situates pedagogy as the generative core of ethical and inclusive artificial intelligence. Design/methodology/approach: A systematic literature review was conducted following PRISMA 2020 guidelines across seven databases (Scopus, Web of Science, IEEE Xplore, ScienceDirect, SpringerLink, Wiley, and Google Scholar) covering publications from 2014 to May 2024. Using the PICo framework for scoping and CIMO logic for synthesis, 127 peer-reviewed studies were analyzed to identify mechanisms linking ethical AI to inclusive educational outcomes. Findings: Four recurrent mechanisms were synthesized from the literature: (1) pedagogical capacity-building , (2) technical process control , (3) regulatory accountability , and (4) participatory legitimation . Integrating these mechanisms, the study proposes the Pedagogically-Centered AI Governance (PCAG) model and the Sustainable Educational Governance Framework (SEG-F) . Together, they conceptualize AI not merely as a technological tool but as an educational infrastructure that should be ethically and pedagogically governed to promote equity, inclusion, and epistemic justice. Practical implications: PCAG offers a multi-layered governance architecture and a six-item operational checklist for policymakers, educators, and designers to translate AI ethics into curriculum development, teacher training, and institutional decision-making. Originality/value: This review advances AI-in-education scholarship by combining PRISMA + PICo + CIMO into a mechanism-oriented synthesis and by centering pedagogy as the key lever for converting global AI ethics principles into contextually just and inclusive educational practices. Ethical AI Inclusive education Pedagogically-Centered AI Governance (PCAG) Systematic review Educational ethics AI governance. 1. Introduction 1.1. Context and Significance Artificial Intelligence (AI) has evolved from a specialized technological tool into a pervasive epistemic infrastructure within education, fundamentally reshaping how knowledge is produced, governed, and distributed (Floridi, 2014 ; Selwyn, 2021b ; Zawacki-Richter et al., 2019 ). Across educational systems, AI-powered technologies—ranging from adaptive learning platforms and predictive analytics to automated essay scoring and intelligent tutoring systems—are redefining pedagogical practices, administrative decision-making, and the very nature of learning itself (Baker, 2016 ; Holmes et al., 2022 ). According to UNESCO ( 2021b ), AI adoption in education is accelerating exponentially; yet, discourses on ethics and inclusivity lag far behind implementation (Shah, 2023 ). This growing disparity underscores an urgent need to reconceptualize AI not merely as a technical add-on but as a strategic educational infrastructure for developing sustainable, equitable, and just learning ecosystems (Williamson & Eynon, 2020 ). Within this transformation, education faces what may be termed a “Janus-faced” reality of AI—one side promising innovation and access, the other exposing deep ethical and social risks. 1.2. Problem and Gap AI offers immense potential to personalize learning, automate administrative processes, and expand access to high-quality education, thereby supporting the achievement of Sustainable Development Goal 4 (Luckin, 2018 ; UNESCO, 2019 ). However, its deployment also raises serious ethical and social challenges, including algorithmic bias that reinforces inequality, digital exclusion that marginalizes vulnerable learners, and privacy risks that threaten autonomy and trust (Bender et al., 2021 ; Eynon, 2022 ; Kritt & Winegar, 2007 ). As Selwyn ( 2021a ) notes, AI-driven personalization may inadvertently deepen the digital divide if not governed by inclusive and justice-oriented principles. The tension between efficiency and equity renders AI a double-edged instrument: capable of creating shared educational value, but only within a governance framework that prioritizes fairness, accountability, and social good (Brynjolfsson, E., & McAfee, 2017 ). Despite a growing body of research on AI ethics and governance, three interconnected gaps persist, limiting the development of actionable frameworks for educational institutions: The Conceptual–Pedagogical Gap : Existing studies articulate normative principles—fairness, transparency, accountability—but rarely connect them to mechanisms of pedagogical value creation and student capability development (Leslie, 2019 ; Whittlestone et al., 2019 ). Ethics remains a conceptual aspiration rather than a pedagogical driver. The Empirical–Contextual Gap : Research across technical AI ethics, educational theory, and institutional governance remains fragmented, preventing holistic understanding of how fairness, inclusivity, and sustainability can co-exist in specific learning contexts (Eynon, 2022 ; Kewalramani et al., 2021 ). The Methodological–Substantive Gap : Most reviews remain descriptive. There is still no comprehensive synthesis explaining the mechanisms —the how and why—by which ethical and inclusive AI interventions yield tangible outcomes such as equity, capability-building, and sustainable learning. No existing study has yet integrated normative principles ( the what ), empirical evidence ( the where and who ), and governance architectures ( the how ) into a unified education-centered framework of global scope. 1.3. Theoretical Positioning This study positions itself at the intersection of AI ethics, educational governance , and pedagogical theory. It draws from Floridi’s ( 2014 ) notion of AI as an epistemic infrastructure and from educational ethics frameworks that view pedagogy as a moral practice (Eynon, 2022 ; Selwyn, 2021b ). Building on the analytical logics of PRISMA, PICo, and CIMO , the study advances a mechanism-oriented synthesis that explains how ethical AI can be pedagogically operationalized to promote inclusive and sustainable education. Prior systematic reviews have primarily offered thematic mappings or typologies of ethical issues (Fjeld et al., 2020 ; Jobin et al., 2019 ; UNESCO, 2021a ; Zawacki-Richter et al., 2019 ). However, these reviews are largely principle-driven—identifying ethical domains—but stop short of revealing the causal and pedagogical pathways by which ethics shape educational practice. The present study moves beyond principle enumeration toward an integrative understanding of mechanisms that connect ethical AI governance to inclusive pedagogical outcomes. Unlike prior reviews that remained descriptive (e.g., Jobin et al., 2019 ; Fjeld et al., 2020 ), this study elucidates the pedagogical mechanisms that translate ethics into actionable governance, bridging theoretical norms with institutional and classroom practice. 1.4. Research Questions and Objectives Guided by the PICo framework—Population (educational institutions adopting AI), Interest (ethical and inclusive AI), and Context (digital-age governance for sustainable and equitable education)—this review addresses three core research questions: RQ1 : How does the international literature frame the relationship between AI ethics and inclusiveness in developing sustainable knowledge societies? RQ2 : What mechanisms explain AI’s contribution to value creation, governance, and educational equity in the digital age? RQ3 : What are the main opportunities and challenges arising from the application of ethical and inclusive AI across disciplines, and how can these insights inform a global governance framework? The overarching objective is to uncover and systematize the mechanisms that translate ethical and inclusive AI principles into pedagogical, institutional, and policy-level practices. 1.5. Contribution and Novelty Through this integration, it provides the first comprehensive, education-centered synthesis that: Explains how and why ethical AI mechanisms operate across pedagogical, institutional, and governance levels. Proposes two theoretical models—the Pedagogically-Centered AI Governance (PCAG) model and the Sustainable Educational Governance Framework (SEG-F) —that together conceptualize pedagogy as the ethical core of AI governance. Bridges abstract ethical principles with actionable governance interventions for educators, policymakers, and designers, contributing to the creation of equitable and sustainable knowledge societies. In doing so, the research advances AI-in-education scholarship from principle-driven description toward mechanism-based understanding, positioning pedagogy as the ethical infrastructure through which AI’s societal value in education is realized. 2. Theoretical Anchor 2.1. AI as Epistemic Infrastructure: Reshaping the Foundations of Education The concept of knowledge societies (Anheier & Hoelscher, 2015 ; Stehr, 2017 ; UNESCO, 2005 ) posits that knowledge is the central resource for sustainable development. In the digital age, Artificial Intelligence (AI) has emerged as a transformative epistemic infrastructure —a foundational system that fundamentally alters how knowledge is produced, curated, validated, and disseminated (Floridi, 2014 ; Williamson, 2024 ). Within education, this is not a distant future but a present reality. AI-driven systems—from intelligent tutoring systems and plagiarism detectors to predictive analytics and automated curriculum generators—are actively constituting what counts as knowledge, who is considered a knower, and what forms of learning are valued (Knox, 2020 ; Selwyn, 2021a ). Floridi's (2014) concept of the "infosphere" is thus directly operationalized in the classroom and the university, where human-AI interactions are reshaping the educational knowledge ecology. Consequently, for education to remain a cornerstone of a sustainable knowledge society, it must engage with AI not only for its technical efficiency but also, and more importantly, for its profound social, ethical, and epistemic implications, demanding a core focus on inclusiveness and justice (Holmes et al., 2022 ; UNESCO, 2023 ). 2.2. Ethical AI and Inclusivity: From Abstract Principles to Educational Practice The discourse on AI ethics has produced key principles such as beneficence, non-maleficence, autonomy, justice, and explicability (Floridi et al., 2018 ). However, critical scholarship reveals that these principles often remain abstract, failing to address the situated realities of educational practice (Bozkurt, 2023 ; Crawford, 2021 ). In an educational context, the principle of justice must be translated into a proactive commitment to educational equity , ensuring that AI systems do not perpetuate historical biases against marginalized student populations (Baker & Hawn, 2022 ). Inclusivity demands more than just access; it requires the active participation of diverse educational stakeholders (teachers, students, parents) in the design and governance of AI systems to avoid structural bias and ensure the benefits of AI are distributed fairly (Jobin et al., 2019 ). The OECD ( 2022 ) rightly emphasizes that AI ethics must be operationalized through concrete governance and regulation, moving beyond mere normative statements to actionable frameworks for schools and universities. 2.3. Educational Governance and Pedagogical Value Creation From an educational leadership perspective, AI governance represents a new frontier for responsible innovation (Von Schomberg, 2013 ). It is not merely a defensive risk-management exercise but a strategic opportunity to drive pedagogically-sound value creation . Building on Porter & Kramer's (2011) concept of shared value, ethical and inclusive AI in education expands this horizon by ensuring that technological innovation is intrinsically linked to core educational missions: enhancing human learning, fostering critical thinking, and promoting well-being (Zhang & Aslan, 2021 ). AI governance, therefore, becomes the mechanism to ensure that algorithmic decisions and data-driven practices do not reinforce exclusion or reduce education to metrics, but instead empower educators and promote sustainable, equitable growth for all learners (Williamson et al., 2020 ). 2.4. Educational Equity and Capability Building in the AI Era In education, AI is often framed as a catalyst for personalized learning and universal access (Luckin, 2018 ). However, this techno-optimistic view is critically challenged by scholars who warn of "technological solutionism"—the risk of applying AI to complex socio-pedagogical problems without addressing underlying issues of power, context, and inequality (Kritt & Winegar, 2007 ; Selwyn, 2019 ). True educational equity in the digital era, as framed by UNESCO ( 2021b ), requires moving beyond access to technology and towards the development of capabilities —the freedom to achieve valued forms of learning and agency (Sen, 2017 ). This entails integrating an equity framework that ensures AI-based education cultivates these capabilities and is aligned with the humanistic goals of SDG 4, rather than undermining them. Through CIMO, ethics is not merely a normative domain but a causal structure that generates sustainability. 2.5. Integrating CIMO as an Analytical Lens for Educational Inquiry To unify these interconnected theoretical strands into a coherent analytical tool, this study employs the CIMO (Context-Intervention-Mechanism-Outcome) logic (Denyer et al., 2008 ), specifically tailored for educational research. Context (C): The global educational landscape characterized by datafication, algorithmic bias, digital exclusion, and ethical risks to student autonomy and equity. Intervention (I): The application of AI in educational settings, governed by ethical and inclusive principles. Mechanism (M): The causal forces that explain how the intervention leads to outcomes. This study focuses on educational governance processes, pedagogical value creation, and equity-enhancing capabilities as the core mechanisms. Outcome (O): The realization of sustainable, just, and equitable educational environments as the bedrock of sustainable knowledge societies. The CIMO framework facilitates the integration of our multidisciplinary theoretical anchor—knowledge society theory (Stehr, 2017 ; UNESCO, 2019 ), AI as epistemic infrastructure (Floridi et al., 2018 ; Williamson et al., 2020 ), critical AI ethics (Baker & Hawn, 2022 ; Bozkurt, 2023 ), responsible innovation & value creation (Porter & Kramer, 2011 ; Von Schomberg, 2013 ), and educational equity as capability building (Luckin, 2018 ; Selwyn, 2021a ; Sen, 2017 ). This synthesis results in a visionary theoretical framework that not only maps the challenges but also provides a causal logic for how ethical and inclusive AI governance can serve as the foundation for sustainable education. The integration of these conceptual axes is visualized in Fig. 1 (see supplementary appendix A) , which illustrates the dynamic connections between knowledge societies, AI as epistemic infrastructure in education, inclusive ethics, educational governance for pedagogical value, and equity as capability building, all synthesized through the CIMO framework towards the ultimate outcome of sustainable educational ecosystems. 3. Methodology (Revised Q1-ready version) 3.1. Study Design This study adopts a systematic literature review (SLR) design to ensure methodological rigor, transparency, and reproducibility in examining the intersection of AI ethics, inclusivity , and sustainability in education (see supplementary Appendix A Figs. 2 and 3 ). The review integrates three complementary frameworks: PICo (Population–Interest–Context) to define the conceptual and contextual boundaries of the research; PRISMA 2020 (Page et al., 2021 ) to structure the literature search, screening, and reporting phases; and CIMO logic (Denyer et al., 2008 ) to synthesize causal mechanisms linking ethical and inclusive AI interventions to sustainable educational outcomes. The analysis followed a thematic synthesis approach (Braun & Clarke, 2006 ), supported by NVivo 14 software to ensure coding consistency. Methodological rigor was reinforced through intercoder reliability testing (O’Connor & Joffe, 2020 ), data triangulation, and critical appraisal using established quality assessment tools. This multi-layered approach enables the review not only to aggregate existing evidence but also to uncover the underlying mechanisms through which ethical and inclusive AI can be operationalized across diverse educational contexts. 3.2. Research Framework: PICo The PICo framework (Methley et al., 2014 ) was employed to clarify the scope of the research questions and ensure conceptual coherence. Unlike the quantitative-oriented PICO model, PICo is suitable for conceptual, normative, and qualitative inquiries, allowing for exploration of ethical and social dimensions beyond technical performance. Population (P) : Public, private, and educational organizations implementing AI systems. These entities represent the operational interface where ethical principles are translated into policies and pedagogical practices. Interest (I) : The focus is on ethical principles and inclusivity to identify value systems embedded within AI’s design, deployment, and governance processes. Context (Co) : The digital era, characterized by algorithmic decision-making and rapid technological disruption, provides a critical backdrop for examining governance, sustainability, and equity. This framing ensures that the study addresses the AI–ethics–inclusivity–sustainability nexus , rather than restricting attention to technical or algorithmic dimensions alone. 3.3. Literature Search Strategy: PRISMA 2020 Following PRISMA 2020 guidelines (Page et al., 2021 ), the literature search prioritized transparency, replicability, and accountability. The search covered seven major databases— Scopus, Web of Science Core Collection, IEEE Xplore, ScienceDirect, SpringerLink, Wiley Online Library , and grey literature from international policy organizations (UNESCO, OECD, and the World Bank). This multidisciplinary approach mitigated publication bias and captured diverse epistemic perspectives (Tranfield et al., 2003 ). A systematic search string was constructed using Boolean and proximity operators aligned with each database’s syntax: (“artificial intelligence” OR “AI” OR “machine learning”) AND (“ethic*” OR “responsible AI” OR “responsible artificial intelligence”) AND (“inclusiv*” OR “equit*” OR “fair*” OR “accessibility”) AND (“governance” OR “policy” OR “value creation” OR “education” OR “sustainable development”) . The search period was January 2014 to May 2024 , capturing the decade of rapid deep learning advancement and the rise of global ethical-AI debates. All identified records were imported into Zotero for deduplication and then screened using Rayyan AI , which enhances efficiency and reduces reviewer bias. Screening was conducted independently by two reviewers, with disagreements resolved through consensus discussion. Intercoder agreement was computed using Cohen’s κ = 0.78 (title/abstract phase) and κ = 0.72 (full-text phase), indicating substantial reliability. The PRISMA flow (Fig. 3 in supplementary Appendix A) summarizes the four-stage process: Identification : 6,785 records retrieved. Deduplication : 1,322 duplicates removed. Screening : 5,463 titles and abstracts assessed; 5,153 excluded as irrelevant. Eligibility : 310 full-text articles reviewed; 183 excluded for specific reasons (see below). Inclusion : 127 studies retained for final synthesis and thematic analysis. Reasons for exclusion included: (n = 98) Irrelevant focus — AI ethics or inclusivity discussed only peripherally. (n = 45) Wrong publication type — non-peer-reviewed materials (magazines, theses, or informal reports). (n = 25) Context mismatch — studies unrelated to education or transferable governance contexts. (n = 15) Lack of thematic integration — addressed ethics or inclusion but not their interconnection in governance or pedagogy. This systematic reduction—from 6,785 initial records to 127 eligible studies—ensures that the synthesis rests exclusively on high-quality, directly relevant literature addressing the intertwined themes of AI ethics, inclusivity, and governance. 3.4. Data Analysis: Thematic Synthesis The thematic analysis followed the six-stage procedure by Braun & Clarke ( 2006 ), enabling systematic pattern identification across heterogeneous studies. NVivo 14 was employed to manage codes, maintain traceability, and visualize emerging themes. The analytical process entailed: (1) Familiarization with data; (2) Generation of initial codes; (3) Searching for themes; (4) Reviewing and refining themes; (5) Defining and naming themes; and (6) Producing the thematic map and synthesis. Reliability was verified through intercoder testing (O’Connor & Joffe, 2020 ) and peer debriefing among the authors. The thematic synthesis identified four overarching mechanisms that link ethical and inclusive AI practices with sustainable educational outcomes, later elaborated through the CIMO framework. 3.5. Synthesis Framework: CIMO Logic To move beyond descriptive mapping toward explanatory synthesis , thematic results were structured using the CIMO framework (Denyer et al., 2008 ; Rousseau, 2006 ): Context (C) : The global digital ecosystem, characterized by algorithmic bias, inequity, and governance challenges. Intervention (I) : Implementation of AI systems guided by ethical and inclusive principles. Mechanism (M) : Governance processes that enable fairness, accountability, transparency, and value creation. Outcome (O) : Sustainable knowledge societies supported by ethical governance and inclusive education. This structure enables identification of causal mechanisms linking ethical AI governance to equitable educational outcomes. A sensitivity analysis (Popay et al., 2006 ; Tanskanen et al., 2017 ) was conducted to test robustness by: Re-running the synthesis after adjusting inclusion/exclusion thresholds; and Re-examining dominant themes after excluding medium-quality studies. Results showed minimal variation, confirming analytical stability. Hence, the CIMO logic not only guided synthesis but also enhanced the explanatory validity of findings regarding how ethical and inclusive AI contributes to sustainable and just learning ecosystems. 3.6. Rigor, Validity, and Reproducibility To ensure research integrity, the study adhered to best practices for SLRs in social sciences and educational technology: Reliability : Coding reliability established through intercoder agreement (Cohen’s κ values reported above). Validity : Triangulation across diverse data sources—academic articles, organizational reports, and international policy documents. Critical Appraisal : Quality assessed using the Joanna Briggs Institute (JBI) tool for qualitative/theoretical studies and the Mixed Methods Appraisal Tool (MMAT) for empirical studies (see supplementary Appendix B: Matrix C & C1). Transparency : Every PRISMA 2020 stage (identification, screening, eligibility, inclusion) explicitly documented, including the list of excluded studies and justifications (See Supplentary Appendix A: Matrix C1). Reproducibility :All matrix sets (processed data), searches, inclusion/exclusion criteria, coding protocols, and synthesis procedures are archived and made publicly accessible (supplementary Appendix A Matrix A - F) to allow for replication (Tranfield et al., 2003 ). This rigorous methodological protocol ensures that the review provides both empirical credibility and conceptual depth , positioning its findings as a reliable foundation for advancing ethical and inclusive AI governance in global education. 4. Results 4.1. Characteristics of the Included Studies A total of 127 studies met the inclusion criteria (see Supplementary Appendix B, Matrix C2). The corpus demonstrates strong methodological diversity, global reach, and consistent quality, forming a robust evidence base for synthesis. As shown in Matrix A (supplementary Appendix B), conceptual and normative works dominate the field (n = 38; 30%). These studies—comprising philosophical essays, normative frameworks, and ethical commentaries—reflect an evolving scholarly phase in which AI ethics is still establishing its theoretical foundations (Buhmann & Fieseler, 2021 ; D. Schiff et al., 2021 ). The emphasis remains on defining principles and governance ideals rather than testing empirically grounded hypotheses. Table A2 shows a strong international distribution: 44 studies (35%) adopt a global or transnational focus, complemented by substantial contributions from the United States (n = 16; 13%), the United Kingdom (n = 8; 6%), and Australia (n = 8; 6%). Importantly, representation from the Global South is emerging through studies from Latin America (Gutierrez et al., 2022 ; Gutierrez Y Restrepo & Floris, 2022), Brazil (Cantarini, 2022 ), India (Sharma et al., 2022 ), and Africa (Spiegel et al., 2021 ), signaling a gradual diversification of epistemic voices beyond traditional Western-centric narratives. As reported in Matric C & C1, quality assessment used the JBI checklists (n = 115; 82.1%) for normative and theoretical works and the MMAT (n = 25; 17.9%) for empirical studies. This dual-tool strategy ensured comparability across methodological traditions and maintained evaluative rigor. The overall methodological quality is high: 72.1% of studies scored ≥ 70 on a 100-point scale (average ≈ 73.5%). No study was excluded for low quality, though 23 (18.1%) scored below 65%. Mattrix D indicates strong thematic interconnectedness. Few papers address a single research question (RQ1 only: n = 9; RQ2 only: n = 7; RQ3 only: n = 3). Most contribute to multiple questions—particularly RQ2 & RQ3 (n = 41; 29.3%) and RQ1 & RQ3 (n = 37; 26.4%)—underscoring the multidisciplinary character of the field and the entanglement between ethical theory, governance mechanisms, and educational practice. Overall, these characteristics confirm a methodologically sound and globally distributed evidence base , enabling reliable synthesis of mechanisms linking AI ethics, inclusivity, and sustainability in education. 4.2. Framing the Relationship between AI Ethics and Inclusivity (RQ1) Across the international literature, AI ethics and inclusivity are framed not as linear cause-effect phenomena but as interdependent socio-technical constructs underpinning sustainable knowledge societies . Four dominant framings emerge ( Table 2 ) , each addressing a distinct dimension of the ethical–inclusive AI nexus. Critical–Decolonial Framing : Studies such as Cantarini ( 2022 ), Casillo et al. ( 2024 ), Sijing & Lan ( 2018 ), and Xu ( 2020 ) highlight epistemic injustice and data colonialism as structural barriers to inclusivity. Here, AI ethics is conceptualized as a decolonial project requiring contextualized, locally grounded, and sovereignty-respecting AI systems. Legal–Institutional Framing : Ess et al. ( 2018 ), Price & Cohen ( 2019 ), Smith et al. (2020), and Su ( 2022 ) view inclusivity as an outcome of enforceable legal and institutional safeguards—such as the right to explanation and data-protection mechanisms (e.g., GDPR)—that embed accountability and human oversight into AI governance. Capability–Pedagogical Framing (The Educational Core) : Educational research (Bozkurt, 2023 ; Calo, 2018; Corredor García et al., 2021 ; Rocco et al., 2022 ) frames AI ethics as a learnable capability cultivated through curriculum integration, teacher training, and critical AI literacy. Inclusivity thus becomes a pedagogical outcome rather than a compliance requirement, establishing ethics as a core educational competence. Value-Based Governance Framing : Gorur et al. (2020), Keeble & Blatchly-Lewis ( 2022 ), and Tsafack Chetsa (2021) link ethics and inclusivity to shared-value creation and ecological sustainability, arguing that ethical governance generates inclusivity, which in turn reinforces social legitimacy and long-term system viability. Synthesis . Together, these framings demonstrate a shift from moral abstraction toward practical operationalization. The pedagogical-capability framing emerges as the translational core —the pathway through which ethical principles become lived educational realities. Other framings provide the structural and critical scaffolding that enables pedagogy to function as the ethical engine of sustainable learning. Table 2 Framings of Ethical and Inclusive AI Framing Definitional Focus n (studies) Key References Critical–Decolonial / Structural Challenges epistemic injustice and data colonialism; emphasizes local context, plural epistemologies, and data sovereignty. 18 Cantarini ( 2022 ); Casillo et al. ( 2024 ); Erman & Furendal ( 2022 ) Legal–Institutional Links AI ethics with law, regulation, and accountability frameworks ensuring transparency and human oversight. 34 Ess et al. ( 2018 ); Price & Cohen ( 2019 ); Su ( 2022 ) Capability–Pedagogical Frames ethics as social capacity; focuses on education, AI literacy, and empowerment for inclusive participation. 28 Corredor García et al. ( 2021 ); Rocco et al. ( 2022 ); Bogina et al. ( 2022 ) Value-Based Sustainability Connects ethical AI with shared value creation and ecological responsibility in organizational contexts. 10 Keeble & Blatchly-Lewis ( 2022 ); Sharma et al. ( 2022 ) 4.3. Mechanisms Linking AI to Value Creation, Governance, and Educational Equity (RQ2) Applying the CIMO synthesis revealed four classes of mechanisms through which ethical and inclusive AI contributes to equitable and sustainable education. Their relative importance varies by context, but in educational environments the Pedagogical-Capacity Mechanism is consistently dominant. Pedagogical-Capacity Mechanisms : Interventions such as ethics-infused curricula (Corredor García et al., 2021 ; Rocco et al., 2022 ), teacher professional development (Calo, 2018), and AI literacy programs (Selwyn, 2021a ; Walsh et al., 2022) increase ethical awareness and behavioral competence. These mechanisms cultivate the human capabilities essential for equitable participation in digital societies. Technical-Process Controls : Mechanisms including fairness metrics (Paic, 2019 ; Weber, 2019), explainable-AI techniques (Castelli et al., 2024 ), and audit trails (Weber-Lewerenz, 2021 ) enhance transparency and reduce algorithmic bias. Such controls build trust in educational technologies, enabling responsible adoption across sectors (Ahmad et al., 2021 ; Lee, 2020). Regulatory-Accountability Mechanisms : Legal standards and compliance instruments (Kim et al., 2021; Sapienza, 2022 ; Smith et al., 2020) enforce transparency and rights protection, creating a governance environment conducive to long-term value creation and safeguarding ethical experimentation in education. Participatory-Legitimation Mechanisms : Inclusive design and deliberative approaches (Camaréna, 2021 ; Bhawra et al., 2021; Kästner & Kang, 2020 ) integrate community knowledge, enhance contextual relevance, and strengthen social legitimacy (Gutierrez Y Restrepo & Floris, 2022). Synthesis . Ethical AI’s contribution to value and equity is mediated rather than automatic. Pedagogical capacity functions as the primary driver, while technical, regulatory, and participatory mechanisms provide the enabling infrastructure—trust, accountability, and legitimacy—that make transformative pedagogy possible. Table 3 Mechanisms for Operationalizing Ethical and Inclusive AI Mechanism Functional Description n (studies) Key References Technical–Process Controls Embeds fairness metrics, explainability (XAI), bias audits, and lifecycle data governance into design. 33 Paic ( 2019 ); Castelli et al. ( 2024 ) Regulatory–Accountability Establishes standards, certification, and liability frameworks to enforce transparency and compliance. 29 Price & Cohen ( 2019 ); Sapienza ( 2022 ); Losavio ( 2021 ) Participatory–Legitimation Uses co-design and multi-stakeholder dialogue to ensure inclusion and social legitimacy. 22 Camaréna ( 2021 ); Kästner & Kang ( 2020 ); Dennehy et al. (2021) Pedagogical–Capacity Building Integrates ethics into education, training, and digital literacy to strengthen long-term inclusion. 26 Corredor García et al. ( 2021 ); Rocco et al. ( 2022 ); Khoury et al. ( 2022 ) Summary These mechanisms operate symbiotically—technical and regulatory actions provide structural reliability, while participatory and pedagogical initiatives ensure inclusivity and accountability. Together, they form the operational backbone of ethical AI governance. 4.4. Opportunities, Challenges, and Pathways toward Global Governance (RQ3) Emerging Opportunities. Convergence of principles —a global normative consensus on fairness, transparency, and non-maleficence (Kim et al., 2021; Kiritchenko et al., 2021; Tsafack Chetsa, 2021). Sectoral policy innovation —context-specific ethical guidelines in domains such as healthcare (Khoury et al., 2022 ), automotive (Luetge et al., 2021 ), and smart cities (Keeble & Blatchly-Lewis, 2022 ) offer scalable blueprints. Ethical-enabler technologies —progress in explainable AI, fairness auditing, and differential privacy operationalizes ethics in practice. Global-South capacity building —contextual models from Latin America, India, and Africa (Cantarini, 2022 ; Gutierrez Y Restrepo & Floris, 2022; Sijing & Lan, 2018 ) democratize governance and rebalance epistemic authority. Persistent Challenges. Principle–practice gap —ethical declarations often lack operational translation, resulting in “ethics washing” (D. Schiff et al., 2021 ). Regulatory fragmentation —divergent jurisdictional frameworks create accountability gaps and compliance complexity (Sapienza, 2022 ; Su, 2022 ). Power asymmetry —policy-making remains dominated by Global-North actors and industry (D. S. Schiff et al., 2022 ). Infrastructural bias —historical data imbalances perpetuate inequity (Denton et al., 2021; Quintarelli et al., 2019). Toward a Hybrid Global Governance Framework . The literature collectively advocates a multi-level architecture: Global Normative Layer : universal human-rights-based principles guiding all AI use. Sectoral Operational Layer : domain-specific standards and certification systems (e.g., education, health). Local Implementation Layer : participatory and context-adaptive mechanisms ensuring legitimacy. Supporting Mechanisms : investment in Global-South capacity, open data, and independent audits. Synthesis. Effective ethical-AI governance requires vertical coherence —aligning top-down principles with bottom-up contextualization. Evidence across 69 studies shows that enduring ethical performance depends on integration among these layers (Floridi & Cowls, 2021 ; McKay et al., 2022 ). Table 4 Governance Themes Emerging from the Literature Governance Theme Definitional Focus n (studies) Key References Hybrid Global Framework Aligns universal ethical principles with sectoral modules and local feedback loops. 30 Floridi & Cowls ( 2021 ); Pitt et al. (2019) Sectoral Operationalization Translates principles into sector-specific standards (e.g., education, health, smart cities). 34 McKay et al. ( 2022 ); Luetge et al. ( 2021 ); Keeble & Blatchly-Lewis ( 2022 ) Local / Decolonial Adaptation Embeds governance within local epistemic contexts to protect data sovereignty and justice. 18 Cantarini ( 2022 ); Gutierrez Y Restrepo & Floris (2022) Standards, Audits & Oversight Promotes certification, interoperability, and independent auditing to reinforce trust. 25 Frischknecht-Gruber et al. ( 2022 ); Schiff et al. ( 2022 ) Summary Effective ethical AI governance requires multi-level coherence —balancing top-down standardization with bottom-up contextualization. Alignment across these levels mitigates fragmentation, reinforces accountability, and sustains equity within global AI ecosystems. 4.5. Integrative Synthesis: Themes, Mechanisms, and Strategic Relevance The cross-theme synthesis demonstrates that ethical AI governance is intrinsic to organizational and educational strategy , not an external compliance activity. Four mechanisms— Transparency → Trust, Accountability → Alignment, Inclusivity → Capability, and Sustainability → Value Creation —constitute a cyclical system of strategic knowledge governance: Data → Governance → Innovation → Sustainability → Data Within this loop, ethical intelligence functions as a dynamic capability driving continuous learning, stakeholder trust, and adaptive performance. Ethical AI thus becomes a strategic infrastructure for sustainable knowledge societies. Table 5 Synthesis of Thematic Findings, Mechanisms, and Managerial Implications Theme Core Mechanism (CIMO Synthesis) Empirical Focus / Illustrative Examples Managerial and Strategic Relevance 1. Ethical Foundations and Principles Transparency → Trust Explainable AI tools, open audit trails, and fairness metrics increase institutional legitimacy (Chakraborty et al., 2020 ; Helberger et al., 2020 ). Builds stakeholder confidence, enhances data integrity, and establishes the ethical baseline for responsible innovation. 2. Organizational and Governance Mechanisms Accountability → Alignment Multi-level governance structures, algorithmic audits, and Corporate Digital Responsibility frameworks (Shneiderman, 2020 ; Weber-Lewerenz, 2021 ). Embeds ethical oversight into strategic and operational processes, aligning AI objectives with mission and regulatory standards. 3. Strategic and Operational Integration Inclusivity → Capability Participatory innovation design, ethics-by-design processes, and inclusive data governance (Buhmann & Fieseler, 2021 ; Rajagopal et al., 2024 ). Transforms ethics into organizational capability, fostering innovation readiness, diversity, and resilience. 4. Global and Cross-Sectoral Governance Models Sustainability → Value Creation UNESCO, OECD, and sectoral frameworks in healthcare, education, and agriculture (Ahmad et al., 2021 ; Klerkx & Rose, 2020 ). Links ethical governance to measurable sustainability outcomes, translating social responsibility into long-term competitive advantage. 4.6. Cross-RQ Integration and Conceptual Continuity The findings across RQ1–RQ3 form a coherent continuum: Framings of ethics and inclusivity (RQ1) provide the normative foundation ; Mechanisms of governance and value creation (RQ2) constitute the operational process ; and Opportunities and global governance pathways (RQ3) delineate the strategic horizon . This interconnection validates the study’s core argument: achieving sustainable knowledge societies requires the interplay of ethical framings, operational mechanisms, and multi-level governance coherence. The synthesis directly informs the construction of the Pedagogically-Centered AI Governance (PCAG) and Sustainable Educational Governance Framework (SEG-F) models (see Figs. 4 and 5 in supplemtary A) , which integrate theoretical, empirical, and normative dimensions into a replicable architecture for global education systems. 5. Discussion and Conceptual Synthesis 5.1. Reframing AI Ethics as Pedagogical Infrastructure The findings collectively reframe Artificial Intelligence not merely as a technological add-on but as a pedagogical infrastructure —a foundation that shapes how knowledge is produced, distributed, and governed (Floridi, 2014; Selwyn, 2021b; Holmes et al., 2022). The global literature reveals a transition from normative reflection toward functional operationalization : AI ethics now operates as a constitutive condition for equitable learning ecosystems , rather than as an external constraint (Buhmann & Fieseler, 2021; Eynon, 2022). This reframing aligns with the critical–decolonial and capability–pedagogical framings (RQ1), where inclusivity and fairness are not end goals but pedagogical capabilities to be cultivated (Bozkurt, 2023; Corredor García et al., 2021). Hence, ethics and inclusivity become dynamic processes embedded in curriculum design, teacher education, and AI literacy , directly influencing learners’ epistemic agency and social participation. In this sense, AI functions as a meta-learning system —an epistemic infrastructure that governs how educational value is created and who participates in it (Williamson & Eynon, 2020; Zawacki-Richter et al., 2019). This reconceptualization grounds the Pedagogically-Centered AI Governance (PCAG) model proposed in this study. 5.2. The Pedagogically-Centered AI Governance (PCAG) Model The PCAG model synthesizes results from RQ1 and RQ2 into a multi-mechanistic framework that explains how ethical and inclusive AI generates educational and societal value . It situates pedagogical capacity-building as the central mechanism (CIMO synthesis), supported by three enabling infrastructures: technical-process controls , regulatory-accountability systems , and participatory-legitimation structures (see Table 3). (1) Pedagogical Core At its heart, PCAG emphasizes AI ethics as a learnable and teachable capability . When ethics is integrated into the curriculum, professional training, and digital literacy programs, it fosters collective epistemic resilience (Rocco et al., 2022; Calo, 2018; Walsh et al., 2022). This “pedagogical core” transforms normative principles—fairness, transparency, accountability—into social capabilities that underpin sustainable learning societies (Bozkurt, 2023; Gutierrez Y Restrepo & Floris, 2022). (2) Technical and Regulatory Infrastructure Technical and regulatory mechanisms establish the trust architecture necessary for ethical learning systems. Fairness metrics (Paic, 2019), explainable AI (Castelli et al., 2024), and algorithmic audits (Weber-Lewerenz, 2021) operationalize transparency, while legal frameworks (Smith et al., 2020; Sapienza, 2022) institutionalize accountability. Together, they create a stable environment where educational actors can engage safely with AI, bridging the principle–practice gap that often undermines ethical implementation (D. Schiff et al., 2021). (3) Participatory and Decolonial Anchors The participatory mechanisms identified (Camaréna, 2021; Bhawra et al., 2021; Kästner & Kang, 2020) ensure that AI development in education remains locally relevant, culturally responsive, and epistemically just (Cantarini, 2022; Casillo et al., 2024). By embedding co-design and local deliberation, PCAG counteracts the Global North bias in AI governance (D. S. Schiff et al., 2022) and affirms the agency of educators and learners as co-creators of ethical technology. (4) Systemic Integration These three subsystems—pedagogical, technical–regulatory, and participatory—form an interdependent architecture. Pedagogical capacity provides the agency layer , technical and regulatory structures supply the integrity layer , and participatory mechanisms contribute the legitimacy layer . Their interaction constitutes the ethical–inclusive AI ecosystem capable of generating trust, equity, and sustainable innovation (Ahmad et al., 2021; Tsafack Chetsa, 2021). Figure 4 visualizes this integrative model: ethics operates through pedagogy, is stabilized by governance, and validated through participation. 5.3. Integrated Framework: From Pedagogy to Global Governance The integration of the Pedagogically-Centered AI Governance (PCAG) and the Sustainable Educational Governance Framework (SEG-F) represents a holistic architecture that connects micro-level educational mechanisms with macro-level global governance systems. This integration explains how ethical and inclusive AI transitions from classroom pedagogy to global sustainability structures. At the pedagogical level , PCAG situates ethics as a learnable capability embedded within educational design, teacher training, and AI literacy programs (Bozkurt, 2023; Corredor García et al., 2021; Rocco et al., 2022). Here, ethics functions not as a regulatory constraint but as a transformative learning process that empowers learners and educators to critically engage with technology. This process forms the capability-building foundation for sustainable digital societies (Calo, 2018; Walsh et al., 2022). At the institutional and regulatory level , the SEG-F framework extends these pedagogical foundations into broader governance coherence. It aligns universal ethical principles (fairness, accountability, transparency) with sectoral operationalization and local adaptation (Floridi & Cowls, 2021; McKay et al., 2022; UNESCO, 2021). The framework ensures that ethical AI practices in education are supported by technical-process controls (Paic, 2019; Castelli et al., 2024), enforceable standards (Smith et al., 2020; Sapienza, 2022), and participatory feedback loops that sustain contextual legitimacy (Camaréna, 2021; Gutierrez Y Restrepo & Floris, 2022). Together, these two frameworks form a continuum of governance learning systems . PCAG functions as the pedagogical engine that internalizes ethics and inclusivity into learning, while SEG-F provides the structural ecosystem that scales these values through multi-level governance. The integration operationalizes Floridi’s concept of the infosphere —where knowledge, data, and ethics co-evolve within an interconnected digital ecology—and extends Sen’s capability approach by demonstrating how ethical governance expands individuals’ real freedoms to learn, participate, and innovate within AI-mediated environments. This convergence thus reframes AI ethics as both epistemic and systemic infrastructure . Ethical AI in education is no longer confined to compliance documents or curricular modules; it becomes a dynamic architecture that links pedagogy, governance, and sustainability. By embedding human capability and moral intelligence within the digital ecosystem, the integrated PCAG–SEG-F framework provides a blueprint for building ethical knowledge societies that are simultaneously locally grounded and globally coherent. 5.4. Policy and Practice Implications The synthesis has direct implications for educational policymakers, institutional leaders, and AI developers. Policy Integration: Ethical AI must be embedded in education policy and accreditation standards , linking digital transformation with social inclusion (UNESCO, 2023). Ministries of education should integrate AI literacy and data ethics into teacher competency frameworks (Walsh et al., 2022; Corredor García et al., 2021). Institutional Governance: Universities and schools should adopt multi-level governance structures —combining AI ethics committees, transparent data policies, and participatory audit mechanisms—to ensure accountability and contextual responsiveness (Weber-Lewerenz, 2021; Shneiderman, 2020). Technological Co-Design: Developers should move from user-centered to educator-centered and community-centered design, engaging teachers and learners in co-creation processes that embed inclusivity from inception (Camaréna, 2021; Bhawra et al., 2021). Global Collaboration: International organizations (OECD, UNESCO, EU) should facilitate South–South knowledge exchange , supporting capacity building and policy harmonization to address epistemic inequality in AI governance (Cantarini, 2022; Gutierrez et al., 2022). Collectively, these implications advance a practical roadmap for operationalizing ethical AI governance as an educational and societal infrastructure. 5.5. Conceptual Synthesis and Future Directions The synthesis confirms that ethics and inclusivity are not adjuncts but core architectures of sustainable AI-driven education. PCAG provides the micro-foundations—pedagogy, literacy, governance—while SEG-F establishes the macro-structure—principles, standards, and global coherence. Future research should empirically test these models across diverse educational systems, focusing on: longitudinal studies of AI ethics curricula and learning outcomes, comparative analyses of governance structures across regions, and participatory design methodologies that integrate Global South epistemologies. Through such expansion, ethical and inclusive AI can evolve from discourse to infrastructure , ensuring that the next phase of digital transformation is not only intelligent but also just, human-centered, and pedagogically grounded. 6. Conclusion and Recommendations Artificial Intelligence has become a constitutive element of educational transformation—no longer a peripheral tool, but a governing infrastructure that shapes how knowledge, capability, and equity are produced. This review demonstrates that the ethical and inclusive governance of AI is not a secondary concern, but the epistemic foundation of sustainable knowledge societies. Synthesizing evidence from 127 peer-reviewed studies through PRISMA, PICo, and CIMO frameworks, the research conceptualized two interlocking models: The Pedagogically-Centered AI Governance (PCAG) and the Sustainable Educational Governance Framework (SEG-F) . Together, they explain how and why ethical principles translate into educational value creation—by aligning pedagogical capacity, regulatory integrity, and participatory legitimacy. The findings advance three critical insights. First, pedagogy is the operational core of AI ethics: inclusive and critical AI literacy must be cultivated through curriculum, teacher education, and lifelong learning systems. Second, technical and regulatory infrastructures —fairness metrics, algorithmic audits, and accountability frameworks—function as enablers that stabilize ethical practice across institutions. Third, participatory and decolonial approaches ensure that AI systems reflect diverse epistemologies, countering the dominance of Global North narratives in AI governance. These interdependent mechanisms redefine AI ethics from a declarative code into a pedagogical process of empowerment, deliberation, and value creation. From a policy perspective, the study recommends embedding AI ethics into national education strategies and accreditation standards, supported by capacity-building programs and independent audit mechanisms. Global organizations such as UNESCO, OECD, and regional alliances should operationalize the SEG-F by promoting multi-level coherence —linking global normative principles with sectoral adaptation and local implementation. Equally, institutions should establish internal governance architectures that integrate AI ethics committees, open-data charters, and participatory design platforms to ensure transparency, inclusivity, and accountability. Through these mechanisms, ethical AI can evolve from regulatory compliance into a strategic infrastructure for sustainable education . Theoretically, this review contributes to the growing discourse on epistemic sustainability—showing that just as infrastructures determine access to energy or transport, AI infrastructures determine access to knowledge and opportunity . Future research should empirically test the PCAG and SEG-F models across diverse contexts, using longitudinal and comparative methodologies that examine their impact on learning outcomes, institutional governance, and social equity. Methodological innovations—such as mixed-methods AI audits, participatory action research, and design-based policy experiments—will be critical to deepen causal understanding. Ultimately, the ethical future of education depends not on how intelligent our machines become, but on how wisely and inclusively we design the human systems that govern them. Positioning AI governance as a pedagogical process redefines education as the moral infrastructure of the digital age. Declarations Author Contribution The author confirms sole responsibility for the following: study conception and design, data collection, analysis and interpretation of results, and manuscript preparation. The author reviewed and approved the final version of the manuscript. Data Availability Data Availability StatementAll data supporting the conclusions in this article are presented within the article itself and in the accompanying Supplementary Appendix. This appendix includes the search matrix, inclusion/exclusion criteria, and a complete data synthesis table. References Ahmad, M. A., Overman, S., Allen, C., Kumar, V., Teredesai, A., & Eckert, C. (2021). Software as a Medical Device: Regulating AI in Healthcare via Responsible AI . 4023–4024. https://doi.org/10.1145/3447548.3470823 Anheier, H. K., & Hoelscher, M. (2015). The 2005 UNESCO Convention and Civil Society: An Initial Assessment. In Globalization, Culture, and Development (pp. 182–202). Palgrave Macmillan UK. https://doi.org/10.1057/9781137397638_13 Baker, R. S. (2016). Stupid Tutoring Systems, Intelligent Humans. International Journal of Artificial Intelligence in Education , 26 (2), 600–614. https://doi.org/10.1007/s40593-016-0105-0 Baker, R. S., & Hawn, A. (2022). Algorithmic Bias in Education. International Journal of Artificial Intelligence in Education , 32 (4), 1052–1092. https://doi.org/10.1007/s40593-021-00285-9 Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the dangers of stochastic parrots: Can language models be too big? FAccT 2021 - Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency , 610–623. https://doi.org/10.1145/3442188.3445922 Bogina, V., Hartman, A., Kuflik, T., & Shulner-Tal, A. (2022). Educating Software and AI Stakeholders About Algorithmic Fairness, Accountability, Transparency and Ethics. International Journal of Artificial Intelligence in Education , 32 (3), 808–833. https://doi.org/10.1007/s40593-021-00248-0 Bozkurt, A. (2023). Generative artificial intelligence (AI) powered conversational educational agents: The inevitable paradigm shift. Asian Journal of Distance Education , 18 (1), 2023. https://www.asianjde.com/ojs/index.php/AsianJDE/article/view/718 Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology , 3 (2), 77–101. https://doi.org/10.1191/1478088706qp063oa Brynjolfsson, E., & McAfee, A. (2017). Artificial intelligence, for real . Harvard Business Review Press. Buhmann, A., & Fieseler, C. (2021). Towards a deliberative framework for responsible innovation in artificial intelligence. Technology in Society , 64 . https://doi.org/10.1016/j.techsoc.2020.101475 Camaréna, S. (2021). Engaging with artificial intelligence (AI) with a bottom-up approach for the purpose of sustainability: Victorian farmers market association, Melbourne Australia. Sustainability (Switzerland) , 13 (16). https://doi.org/10.3390/su13169314 Cantarini, P. (2022). GOVERNANCE OF AI ALGORITHMS-AN APPROACH FROM EPISTEMOLOGIES OF THE SOUTH. Revista Juridica , 5 (72), 393–423. https://doi.org/10.26668/revistajur.2316-753X.v5i72.6286 Casillo, M., Cecere, L., Dembele, S. P., Lorusso, A., Santaniello, D., & Valentino, C. (2024). The Metaverse and Revolutionary Perspectives for the Smart Cities of the Future. In X.-S. Yang, S. Sherratt, N. Dey, & A. Joshi (Eds.), Lecture Notes in Networks and Systems: Vol. 1004 LNNS (pp. 215–225). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-97-3305-7_17 Castelli, T., Mocenni, C., & Maria Dimitri, G. (2024). A machine learning approach to assess Sustainable Development Goals food performances: The Italian case. PLOS ONE , 19 (1 January). https://doi.org/10.1371/journal.pone.0296465 Chakraborty, J., Majumder, S., Yu, Z., & Menzies, T. (2020). Fairway: A way to build fair ML software. In P. Devanbu, M. Cohen, & T. Zimmermann (Eds.), ESEC/FSE 2020 - Proceedings of the 28th ACM Joint Meeting European Software Engineering Conference and Symposium on the Foundations of Software Engineering (pp. 654–665). Association for Computing Machinery, Inc. https://doi.org/10.1145/3368089.3409697 Corredor García, M. P., Quintero Rengifo, E. F., Ariza Ruiz, E. D., & Piñeros, M. I. A. (2021). Methodological Tools Design to Teach Ethics in Accounting According to IFAC IES 4. An Approximation to the Giving Voice to Values (GVV) Methodology. In M. Botto-Tobar, H. Cruz, & A. Díaz Cadena (Eds.), Advances in Intelligent Systems and Computing: Vol. 1327 AISC (pp. 17–28). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-68083-1_2 Crawford, K. (2021). Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. In Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence . https://doi.org/10.56315/pscf3-22crawford Dennehy, D., Griva, A., Pouloudi, N., Dwivedi, Y. K., Dwivedi, Y. K., Pappas, I., Pappas, I., & Mantymaki, M. (Eds.). (2021). 20th IFIP WG 6.11 Conference on e-Business, e-Services and e-Society, I3E 2021. Lecture Notes in Computer Science , 12896 LNCS . https://www.scopus.com/inward/record.uri?eid=2-s2.0-85115191822&partnerID=40&md5=dd33c6f1dc80f55f3f9baa7b5865bb76 Denyer, D., Tranfield, D., & van Aken, J. E. (2008). Developing Design Propositions through Research Synthesis. Organization Studies , 29 (3), 393–413. https://doi.org/10.1177/0170840607088020 Erman, E., & Furendal, M. (2022). The Global Governance of Artificial Intelligence: Some Normative Concerns. Moral Philosophy and Politics , 9 (2), 267–291. https://doi.org/10.1515/mopp-2020-0046 Ess, C., Kreps, D., Kimppa, K., & Leenen, L. (Eds.). (2018). 13th IFIP TC 9 International Conference on Human Choice and Computers, HCC 2018 Held at the 24th IFIP World Computer Congress, WCC 2018. IFIP Advances in Information and Communication Technology , 537 . https://www.scopus.com/inward/record.uri?eid=2-s2.0-85053888611&partnerID=40&md5=bb8edacf59336970cb7a84e8f049db24 Eynon, R. (2022). DATAFICATION AND THE ROLE OF SCHOOLING: Challenging the status quo. Learning to Live with Datafication: Educational Case Studies and Initiatives from Across the World , 17–34. https://doi.org/10.4324/9781003136842-2 Fjeld, J., Nagy, A., Hilligoss, H., Achten, N., & Srikumar, M. (2020). Principled Artificial Intelligence: Mapping Consensus in Ethical and Rights-Based Approaches to Principles for AI. SSRN Electronic Journal . https://doi.org/10.2139/ssrn.3518482 Floridi, L. (2014). The Fourth Revolution: How the Infosphere is Reshaping Human Reality . OUP Oxford. https://books.google.co.id/books?id=hOF_AwAAQBAJ Floridi, L., & Cowls, J. (2021). A Unified Framework of Five Principles for AI in Society. In Philosophical Studies Series (Vol. 144, pp. 5–17). https://doi.org/10.1007/978-3-030-81907-1_2 Floridi, L., Cowls, J., Beltrametti, M., Chatila, R., Chazerand, P., Dignum, V., Luetge, C., Madelin, R., Pagallo, U., Rossi, F., Schafer, B., Valcke, P., & Vayena, E. (2018). AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations. Minds and Machines , 28 (4), 689–707. https://doi.org/10.1007/s11023-018-9482-5 Frischknecht-Gruber, C. M.-L., Reif, M., & Senn, C. W. (2022). Global Efforts Towards Establishing Safety Directives for Intelligent Systems: Review (M. C. Leva, E. Patelli, L. Podofillini, & S. Wilson (Eds.); pp. 2420–2427). Research Publishing. https://doi.org/10.3850/978-981-18-5183-4_S11-11-472-cd Gutierrez, S. S. M., Perez, S. L., & Munguia, M. G. (2022). Artificial Intelligence in e-Learning Plausible Scenarios in Latin America and New Graduation Competencies. Revista Iberoamericana de Tecnologias Del Aprendizaje , 17 (1), 31–40. https://doi.org/10.1109/RITA.2022.3149833 Gutierrez Y Restrepo, E., & Floris, C. (2022). Supporting Teachers and Students Through a Smart Integrated System for Truly Inclusive Higher Education . 107–113. https://doi.org/10.1145/3563137.3563173 Helberger, N., Araujo, T., & de Vreese, C. H. (2020). Who is the fairest of them all? Public attitudes and expectations regarding automated decision-making. Computer Law and Security Review , 39 . https://doi.org/10.1016/j.clsr.2020.105456 Holmes, W., Porayska-Pomsta, K., Holstein, K., Sutherland, E., Baker, T., Shum, S. B., Santos, O. C., Rodrigo, M. T., Cukurova, M., Bittencourt, I. I., & Koedinger, K. R. (2022). Ethics of AI in Education: Towards a Community-Wide Framework. International Journal of Artificial Intelligence in Education , 32 (3), 504–526. https://doi.org/10.1007/s40593-021-00239-1 Jeremy Pitt, J., DIaconescu, A., & Ober, J. (2019). Knowledge Management for Democratic Governance of Socio-Technical Systems. In Lecture Notes in Computer Science: Vol. 11300 LNCS (pp. 38–61). Springer Verlag [email protected] . https://doi.org/10.1007/978-3-030-05333-8_4 Jobin, A., Ienca, M., & Vayena, E. (2019). The global landscape of AI ethics guidelines. Nature Machine Intelligence , 1 (9), 389–399. https://doi.org/10.1038/s42256-019-0088-2 Kästner, C., & Kang, E. (2020). Teaching software engineering for AI-enabled systems. Proceedings - International Conference on Software Engineering , 45–48. https://doi.org/10.1145/3377814.3381714 Keeble, J., & Blatchly-Lewis, M. (2022). The responsible city (pp. 139–154). Wiley Blackwell. https://doi.org/10.1002/9781119746294.ch8 Kewalramani, S., Kidman, G., & Palaiologou, I. (2021). Using Artificial Intelligence (AI)-interfaced robotic toys in early childhood settings: a case for children’s inquiry literacy. European Early Childhood Education Research Journal , 29 (5), 652–668. https://doi.org/10.1080/1350293X.2021.1968458 Khoury, P., Srinivasan, R., Kakumanu, S., Ochoa, S., Keswani, A., Sparks, R., & Rider, N. L. (2022). A Framework for Augmented Intelligence in Allergy and Immunology Practice and Research—A Work Group Report of the AAAAI Health Informatics, Technology, and Education Committee. Journal of Allergy and Clinical Immunology: In Practice , 10 (5), 1178–1188. https://doi.org/10.1016/j.jaip.2022.01.047 Klerkx, L., & Rose, D. (2020). Dealing with the game-changing technologies of Agriculture 4.0: How do we manage diversity and responsibility in food system transition pathways? Global Food Security , 24 . https://doi.org/10.1016/j.gfs.2019.100347 Knox, J. (2020). Artificial intelligence and education in China. Learning, Media and Technology , 45 (3), 298–311. https://doi.org/10.1080/17439884.2020.1754236 Kritt, D. W., & Winegar, L. T. (2007). Education and technology : critical perspectives, possible futures . Lexington Books. Leslie, D. (2019). Understanding artificial intelligence ethics and safety . https://doi.org/10.5281/zenodo.3240529 Losavio, M. (2021). Algorithms of Machines and Law: Risks in Pattern Recognition, Machine Learning and Artificial Intelligence for Justice and Fairness. Public Governance, Administration and Finances Law Review , 6 (2), 21–34. https://doi.org/10.53116/pgaflr.2021.2.3 Luckin, R. (2018). Machine Learning and Human Intelligence: The Future of Education for the 21st Century. In UCL IOE Press . UCL IOE Press. https://eric.ed.gov/?id=ED584840 Luetge, C., Poszler, F., Acosta, A. J., Danks, D., Gottehrer, G., Mihet-Popa, L., & Naseer, A. (2021). AI4people: Ethical guidelines for the automotive sector-fundamental requirements and practical recommendations. International Journal of Technoethics , 12 (1), 101–125. https://doi.org/10.4018/IJT.20210101.oa2 McKay, F., Williams, B. J., Prestwich, G., Bansal, D., Hallowell, N., & Treanor, D. (2022). The ethical challenges of artificial intelligence-driven digital pathology. Journal of Pathology: Clinical Research , 8 (3), 209–216. https://doi.org/10.1002/cjp2.263 Methley, A. M., Campbell, S., Chew-Graham, C., McNally, R., & Cheraghi-Sohi, S. (2014). PICO, PICOS and SPIDER: A comparison study of specificity and sensitivity in three search tools for qualitative systematic reviews. BMC Health Services Research , 14 (1). https://doi.org/10.1186/s12913-014-0579-0 O’Connor, C., & Joffe, H. (2020). Intercoder Reliability in Qualitative Research: Debates and Practical Guidelines. International Journal of Qualitative Methods , 19 . https://doi.org/10.1177/1609406919899220 OECD. (2022). OECD, Recommendation of the Council on Artificial Intelligence, OECD/LEGAL/0449 . http://legalinstruments.oecd.org Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., … Moher, D. (2021). The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ , n71. https://doi.org/10.1136/bmj.n71 Paic, A. (2019). Policies for artificial intelligence in science and innovation. Proceedings of Science , 372 . https://www.scopus.com/inward/record.uri?eid=2-s2.0-85101244456&partnerID=40&md5=9a320da7b8f618b5b7a690bd0f6e2fee Popay, J., Roberts, H., Sowden, A., Petticrew, M., Arai, L., Rodgers, M., & et all. (2006). Guidance on the conduct of narrative synthesis in systematic reviews. a product from the a product from the ESRC methods programme. Esrc , version 1 (April 2006), 1–92. https://doi.org/10.13140/2.1.1018.4643 Porter, M. E., & Kramer, M. R. (2011). Harvard Business Riview. Harvard Business Review , 89 (1–2). Price, W. N., & Cohen, I. G. (2019). Privacy in the age of medical big data. Nature Medicine , 25 (1), 37–43. https://doi.org/10.1038/s41591-018-0272-7 Rajagopal, N. K., Anand, M., & Mohanty, S. (2024). Exploring Machine Learning Applications in Human Resources Management: A Comprehensive Review. In Studies in Systems, Decision and Control (Vol. 569, pp. 303–313). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-71649-2_26 Rocco, R., Thomas, A., & Novas-Ferradás, M. (2022). Teaching design for values: Concepts, tools & practices . TU Delft. https://doi.org/10.34641/mg.54 Rousseau, D. M. (2006). Is there Such a thing as “Evidence-Based Management”? Academy of Management Review , 31 (2), 256–269. https://doi.org/10.5465/amr.2006.20208679 Sapienza, S. (2022). Current Trends, Machine Learning, and Food Safety Data Governance. In Law, Governance and Technology Series (Vol. 52, pp. 123–160). Springer Nature. https://doi.org/10.1007/978-3-031-09367-8_4 Schiff, D., Rakova, B., Ayesh, A., Fanti, A., & Lennon, M. (2021). Explaining the Principles to Practices Gap in AI. IEEE Technology and Society Magazine , 40 (2), 81–94. https://doi.org/10.1109/MTS.2021.3056286 Schiff, D. S., Laas, K., Biddle, J. B., & Borenstein, J. (2022). Global AI Ethics Documents: What They Reveal About Motivations, Practices, and Policies. In International Library of Ethics, Law and Technology (Vol. 23, pp. 121–143). Springer Science and Business Media B.V. https://doi.org/10.1007/978-3-030-86201-5_7 Selwyn, N. (2019). Should Robots Replace Teachers? AI and the Future of Education. In John Wiley & Sons, Inc . Selwyn, N. (2021a). Ed-Tech Within Limits: Anticipating educational technology in times of environmental crisis. E-Learning and Digital Media , 18 (5), 496–510. https://doi.org/10.1177/20427530211022951 Selwyn, N. (2021b). Education and technology: Key issues and debates . Bloomsbury Publishing. Sen, A. (2017). Development as Freedom. In The Top 50 Sustainability Books . Oxford University Press. https://doi.org/10.4324/9781351279086-33 Shah, C. (2023). Generative AI and the Future of Information Access. UNESCO , July , 3–3. https://doi.org/10.1145/3583780.3615317 Sharma, S., Rawal, Y. S., Pal, S., & Dani, R. (2022). Fairness, Accountability, Sustainability, Transparency (FAST) of Artificial Intelligence in Terms of Hospitality Industry. In S. Fong, N. Dey, & A. Joshi (Eds.), Lecture Notes in Networks and Systems (Vol. 314, pp. 495–504). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-16-5655-2_48 Shneiderman, B. (2020). Bridging the gap between ethics and practice: Guidelines for reliable, safe, and trustworthy human-centered AI systems. ACM Transactions on Interactive Intelligent Systems , 10 (4). https://doi.org/10.1145/3419764 Sijing, L., & Lan, W. (2018). Artificial intelligence education ethical problems and solutions. 13th International Conference on Computer Science and Education, ICCSE 2018 , 155–158. https://doi.org/10.1109/ICCSE.2018.8468773 Spiegel, J. M., Wilkinson, J., Ehrlich, R., Lockhart, K., Yassi, A., Barker, S., Riera, F., & Kistnasamy, B. (2021). Using artificial intelligence for high-volume identification of silicosis and tuberculosis a bio-ethics approach. Annals of Global Health , 87 (1). https://doi.org/10.5334/aogh.3206 Stehr, N. (2017). Knowledge Societies. In Society & Knowledge (pp. 299–322). Routledge. https://doi.org/10.4324/9781315129884-13 Su, A. (2022). The Promise and Perils of International Human Rights Law for AI Governance. Law, Technology and Humans , 4 (1). https://doi.org/10.5204/LTHJ.2332 Tanskanen, K., Ahola, T., Aminoff, A., Bragge, J., Kaipia, R., & Kauppi, K. (2017). Towards evidence-based management of external resources: Developing design propositions and future research avenues through research synthesis. Research Policy , 46 (6), 1087–1105. https://doi.org/10.1016/j.respol.2017.04.002 Tranfield, D., Denyer, D., & Smart, P. (2003). Towards a Methodology for Developing Evidence‐Informed Management Knowledge by Means of Systematic Review. British Journal of Management , 14 (3), 207–222. https://doi.org/10.1111/1467-8551.00375 UNESCO. (2005). 2005 Convention on Diversity of Cultural Expressions . UNESCO. (2019). Global Education Monitoring Report 2019: Migration, Displacement and Education – Building Bridges, not Walls. Global Education Monitoring Report 2019: Migration, Displacement and Education – Building Bridges, Not Walls . https://doi.org/10.54676/xdzd4287 UNESCO. (2021a). AI and education: guidance for policy-makers. In AI and education: guidance for policy-makers . UNESCO Publishing. https://doi.org/10.54675/pcsp7350 UNESCO. (2021b). Publications (Vol. 17). http://unesdoc.unesco.org/ UNESCO. (2023). Global Education Monitoring Report 2023: Technology in education: A tool on whose terms? In Global Education Monitoring Report 2023: Technology in education: A tool on whose terms? GEM Report UNESCO. https://doi.org/10.54676/uzqv8501 Von Schomberg, R. (2013). A vision of responsible research and innovation. Responsible Innovation: Managing the Responsible Emergence of Science and Innovation in Society (Owen, R., Bessant, J. and Heintz, M., eds). Responsible Innovation , 51–74. Weber-Lewerenz, B. (2021). Corporate digital responsibility (CDR) in construction engineering—ethical guidelines for the application of digital transformation and artificial intelligence (AI) in user practice. SN Applied Sciences , 3 (10). https://doi.org/10.1007/s42452-021-04776-1 Whittlestone, J., Nyrup, R., Alexandrova, A., Dihal, K., & Cave, S. (2019). Ethical and societal implications of algorithms, data, and artificial intelligence: a roadmap for research. London: Nuffield … , 1–59. https://www.nuffieldfoundation.org/wp-content/uploads/2019/02/Ethical-and-Societal-Implications-of-Data-and-AI-report-Nuffield-Foundat.pdf%0Ahttp://www.nuffieldfoundation.org/sites/default/files/files/Ethical-and-Societal-Implications-of-Data-and-AI-repor Williamson, B. (2024). The Social life of AI in Education. International Journal of Artificial Intelligence in Education , 34 (1), 97–104. https://doi.org/10.1007/s40593-023-00342-5 Williamson, B., Bayne, S., & Shay, S. (2020). The datafication of teaching in Higher Education: critical issues and perspectives. Teaching in Higher Education , 25 (4), 351–365. https://doi.org/10.1080/13562517.2020.1748811 Williamson, B., & Eynon, R. (2020). Historical threads, missing links, and future directions in AI in education. Learning, Media and Technology , 45 (3), 223–235. https://doi.org/10.1080/17439884.2020.1798995 Xu, L. (2020). The Dilemma and Countermeasures of AI in Educational Application. ACM International Conference Proceeding Series , 289–294. https://doi.org/10.1145/3445815.3445863 Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education – where are the educators? International Journal of Educational Technology in Higher Education , 16 (1), 39. https://doi.org/10.1186/s41239-019-0171-0 Zhang, K., & Aslan, A. B. (2021). AI technologies for education: Recent research & future directions. Computers and Education: Artificial Intelligence , 2 , 100025. https://doi.org/10.1016/j.caeai.2021.100025 Additional Declarations No competing interests reported. Supplementary Files SupplementaryAppendixAFigure15.docx SupplementaryBMatrixABCC1C2DD1EF.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8230657","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":554402671,"identity":"e127c87d-5811-49d3-b02d-16ba02868eae","order_by":0,"name":"Dwi Mariyono","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA90lEQVRIiWNgGAWjYDCCA4wNQFICxGR8AGWAgARYnJAWZgMitSCYbBLIEji18N1Ibt3wc49FYr9087Fq3h0W9rrtBxg//GCwkMWlRfJGYtvNnmcSiTPnHEu7zXtGInHbmQRmyR4GCWNcWgxuJ7bd4DkgkbjhRo7Zbd42iQSzAwkM0kC/JOLTcvMPVEsxUIu92fkHzL8JabkNs4UZqIVx240ENry2SN5/2HZb5oCE8cwZacmSc9uAfrnxsM2yxwC3X/jOHH92882BOtl+ieSDH9621QEdlnz4xo+KOpwhBgOOSApAMWJAQD0Q2BNWMgpGwSgYBSMWAADXnV8k5zwDpAAAAABJRU5ErkJggg==","orcid":"","institution":"Universitas Islam Malang","correspondingAuthor":true,"prefix":"","firstName":"Dwi","middleName":"","lastName":"Mariyono","suffix":""}],"badges":[],"createdAt":"2025-11-28 13:08:21","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8230657/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8230657/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":98928948,"identity":"2d4f501e-d8b2-4b43-9ff9-dcebc9ef428c","added_by":"auto","created_at":"2025-12-24 08:24:28","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1252522,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryAppendixAFigure15.docx","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/0f29244faa5a5a2a55d6b6d0.docx"},{"id":98928951,"identity":"54df5fc7-707f-4d01-bd3a-8697a8440ff5","added_by":"auto","created_at":"2025-12-24 08:24:28","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":130300,"visible":true,"origin":"","legend":"","description":"","filename":"EthicalandInclusiveAIasEducational.docx","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/6c1d9f76d90b8ce89fc1c8a3.docx"},{"id":98928846,"identity":"3ffd0528-b3f9-4fb8-b687-e926c2c436a4","added_by":"auto","created_at":"2025-12-24 08:24:18","extension":"json","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4127,"visible":true,"origin":"","legend":"","description":"","filename":"343480d694b641008f5499feab7f93e2.json","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/1e1e758f73dda0b67b00b3e9.json"},{"id":98928938,"identity":"706c017b-0df5-42de-b2e5-c9052eacfb58","added_by":"auto","created_at":"2025-12-24 08:24:22","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":335984,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryBMatrixABCC1C2DD1EF.docx","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/4f6c2b6f6b50d2d7ba2d457c.docx"},{"id":98928956,"identity":"4caa7a6a-7b7d-4583-b838-b7d7960b4e29","added_by":"auto","created_at":"2025-12-24 08:24:29","extension":"xml","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":216050,"visible":true,"origin":"","legend":"","description":"","filename":"343480d694b641008f5499feab7f93e21enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/a57d292c1695a091df726f87.xml"},{"id":98928944,"identity":"cb0dbe6d-ca90-449a-babe-8580ca63425d","added_by":"auto","created_at":"2025-12-24 08:24:28","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":705621,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/827abb4b80a1262c63a5f71a.jpeg"},{"id":98928958,"identity":"92c10e1c-01db-4f9a-8dd2-1609c3c61d17","added_by":"auto","created_at":"2025-12-24 08:24:30","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":298087,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/0533adf5f9bd7de417b931a0.png"},{"id":98928849,"identity":"b610b1d4-83af-40b7-a5a3-8ca35ad3733a","added_by":"auto","created_at":"2025-12-24 08:24:18","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":236476,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/875dd4df0e3f068932bc6f5c.png"},{"id":99310288,"identity":"5ca90ba0-8b73-4e76-8bc9-e5ebb607a74d","added_by":"auto","created_at":"2025-12-31 16:12:27","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":454614,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/cd7a8ef61ba6b15f489c3ac3.jpeg"},{"id":98928911,"identity":"60729b24-09a0-4ff5-b3d2-8965f18f1027","added_by":"auto","created_at":"2025-12-24 08:24:20","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":446415,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/6d75f163772e5b66a07e5f21.png"},{"id":98928913,"identity":"2eb0cbe7-cac8-48f6-ad53-2d1d0b6caede","added_by":"auto","created_at":"2025-12-24 08:24:21","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":129848,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/05e183ed7b385c3a3672964e.png"},{"id":98928905,"identity":"6ddc7f2a-0f62-40ca-b894-16b05f573ddc","added_by":"auto","created_at":"2025-12-24 08:24:19","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":91206,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/9b968a9085f36d0732d64e9f.png"},{"id":98928918,"identity":"77fd59fa-3ccd-4adb-a0e4-ea29bebfc19f","added_by":"auto","created_at":"2025-12-24 08:24:22","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":48304,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/0b29568051a6ac067a97b9c2.png"},{"id":98928957,"identity":"dc67999f-7901-4e87-9980-ff5a7772b527","added_by":"auto","created_at":"2025-12-24 08:24:30","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":82024,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/b6063578835f7ebf1ee8b1fc.png"},{"id":99310223,"identity":"f6512482-41ef-4fa4-98fd-a0caea2dc50c","added_by":"auto","created_at":"2025-12-31 16:12:16","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":43439,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/7869e7b3e31bf7a0d7ce6f81.png"},{"id":98928943,"identity":"01ec661c-ba19-491a-91a6-6caca9ee37a8","added_by":"auto","created_at":"2025-12-24 08:24:28","extension":"xml","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":213044,"visible":true,"origin":"","legend":"","description":"","filename":"343480d694b641008f5499feab7f93e21structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/110c9322939f9112663a62f0.xml"},{"id":98928962,"identity":"189202a7-e080-430d-b691-a37c80aa1047","added_by":"auto","created_at":"2025-12-24 08:24:31","extension":"html","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":240974,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/c2bcee5561c21c80c181d6c0.html"},{"id":104961420,"identity":"5e9d5cf0-9949-4ecf-9c72-660d27f6dfa4","added_by":"auto","created_at":"2026-03-19 08:59:03","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2595969,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/64c20a93-c5ca-402d-874c-e0141988fefe.pdf"},{"id":99310786,"identity":"c84ce3a9-1dc3-41a6-85d4-af9c15292c8a","added_by":"auto","created_at":"2025-12-31 16:13:22","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1252522,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryAppendixAFigure15.docx","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/c2f490c8a579b4b9f93e17e1.docx"},{"id":98928915,"identity":"2ec57738-250f-41be-902e-3ecd92836864","added_by":"auto","created_at":"2025-12-24 08:24:21","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":335984,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryBMatrixABCC1C2DD1EF.docx","url":"https://assets-eu.researchsquare.com/files/rs-8230657/v1/91f41e4c9fa14bcf02559232.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Ethical and Inclusive Artificial Intelligence for Sustainable Knowledge Societies: A Global Governance Framework through a Systematic Literature Review","fulltext":[{"header":"1. Introduction","content":"\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003e1.1. Context and Significance\u003c/h2\u003e \u003cp\u003eArtificial Intelligence (AI) has evolved from a specialized technological tool into a pervasive \u003cem\u003eepistemic infrastructure\u003c/em\u003e within education, fundamentally reshaping how knowledge is produced, governed, and distributed (Floridi, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Selwyn, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e; Zawacki-Richter et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Across educational systems, AI-powered technologies\u0026mdash;ranging from adaptive learning platforms and predictive analytics to automated essay scoring and intelligent tutoring systems\u0026mdash;are redefining pedagogical practices, administrative decision-making, and the very nature of learning itself (Baker, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Holmes et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAccording to UNESCO (\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e), AI adoption in education is accelerating exponentially; yet, discourses on ethics and inclusivity lag far behind implementation (Shah, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This growing disparity underscores an urgent need to reconceptualize AI not merely as a technical add-on but as a strategic \u003cem\u003eeducational infrastructure\u003c/em\u003e for developing sustainable, equitable, and just learning ecosystems (Williamson \u0026amp; Eynon, \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Within this transformation, education faces what may be termed a \u0026ldquo;Janus-faced\u0026rdquo; reality of AI\u0026mdash;one side promising innovation and access, the other exposing deep ethical and social risks.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e1.2. Problem and Gap\u003c/h2\u003e \u003cp\u003eAI offers immense potential to personalize learning, automate administrative processes, and expand access to high-quality education, thereby supporting the achievement of Sustainable Development Goal 4 (Luckin, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; UNESCO, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). However, its deployment also raises serious ethical and social challenges, including algorithmic bias that reinforces inequality, digital exclusion that marginalizes vulnerable learners, and privacy risks that threaten autonomy and trust (Bender et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Eynon, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Kritt \u0026amp; Winegar, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). As Selwyn (\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2021a\u003c/span\u003e) notes, AI-driven personalization may inadvertently deepen the digital divide if not governed by inclusive and justice-oriented principles. The tension between efficiency and equity renders AI a double-edged instrument: capable of creating shared educational value, but only within a governance framework that prioritizes fairness, accountability, and social good (Brynjolfsson, E., \u0026amp; McAfee, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite a growing body of research on AI ethics and governance, three \u003cem\u003einterconnected gaps\u003c/em\u003e persist, limiting the development of actionable frameworks for educational institutions:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eThe Conceptual\u0026ndash;Pedagogical Gap\u003c/b\u003e: Existing studies articulate normative principles\u0026mdash;fairness, transparency, accountability\u0026mdash;but rarely connect them to mechanisms of pedagogical value creation and student capability development (Leslie, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Whittlestone et al., \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Ethics remains a conceptual aspiration rather than a pedagogical driver.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eThe Empirical\u0026ndash;Contextual Gap\u003c/b\u003e: Research across technical AI ethics, educational theory, and institutional governance remains fragmented, preventing holistic understanding of how fairness, inclusivity, and sustainability can co-exist in specific learning contexts (Eynon, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Kewalramani et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eThe Methodological\u0026ndash;Substantive Gap\u003c/b\u003e: Most reviews remain descriptive. There is still no comprehensive synthesis explaining the \u003cem\u003emechanisms\u003c/em\u003e\u0026mdash;the how and why\u0026mdash;by which ethical and inclusive AI interventions yield tangible outcomes such as equity, capability-building, and sustainable learning. No existing study has yet integrated normative principles (\u003cem\u003ethe what\u003c/em\u003e), empirical evidence (\u003cem\u003ethe where and who\u003c/em\u003e), and governance architectures (\u003cem\u003ethe how\u003c/em\u003e) into a unified education-centered framework of global scope.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e1.3. Theoretical Positioning\u003c/h2\u003e \u003cp\u003eThis study positions itself at the intersection of \u003cem\u003eAI ethics, educational governance\u003c/em\u003e, and \u003cem\u003epedagogical theory.\u003c/em\u003e It draws from Floridi\u0026rsquo;s (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) notion of AI as an epistemic infrastructure and from educational ethics frameworks that view pedagogy as a moral practice (Eynon, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Selwyn, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e). Building on the analytical logics of \u003cb\u003ePRISMA, PICo, and CIMO\u003c/b\u003e, the study advances a mechanism-oriented synthesis that explains how ethical AI can be pedagogically operationalized to promote inclusive and sustainable education.\u003c/p\u003e \u003cp\u003ePrior systematic reviews have primarily offered thematic mappings or typologies of ethical issues (Fjeld et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Jobin et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; UNESCO, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2021a\u003c/span\u003e; Zawacki-Richter et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). However, these reviews are largely principle-driven\u0026mdash;identifying ethical domains\u0026mdash;but stop short of revealing the \u003cem\u003ecausal and pedagogical pathways\u003c/em\u003e by which ethics shape educational practice. The present study moves beyond principle enumeration toward an integrative understanding of \u003cem\u003emechanisms\u003c/em\u003e that connect ethical AI governance to inclusive pedagogical outcomes. Unlike prior reviews that remained descriptive (e.g., Jobin et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Fjeld et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), this study elucidates the pedagogical mechanisms that translate ethics into actionable governance, bridging theoretical norms with institutional and classroom practice.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e1.4. Research Questions and Objectives\u003c/h2\u003e \u003cp\u003eGuided by the \u003cb\u003ePICo\u003c/b\u003e framework\u0026mdash;Population (educational institutions adopting AI), Interest (ethical and inclusive AI), and Context (digital-age governance for sustainable and equitable education)\u0026mdash;this review addresses three core research questions:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eRQ1\u003c/b\u003e: How does the international literature frame the relationship between AI ethics and inclusiveness in developing sustainable knowledge societies?\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eRQ2\u003c/b\u003e: What mechanisms explain AI\u0026rsquo;s contribution to value creation, governance, and educational equity in the digital age?\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eRQ3\u003c/b\u003e: What are the main opportunities and challenges arising from the application of ethical and inclusive AI across disciplines, and how can these insights inform a global governance framework?\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eThe overarching objective is to uncover and systematize the mechanisms that translate ethical and inclusive AI principles into pedagogical, institutional, and policy-level practices.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e1.5. Contribution and Novelty\u003c/h2\u003e \u003cp\u003eThrough this integration, it provides the first comprehensive, education-centered synthesis that:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eExplains \u003cem\u003ehow and why\u003c/em\u003e ethical AI mechanisms operate across pedagogical, institutional, and governance levels.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eProposes two theoretical models\u0026mdash;the \u003cb\u003ePedagogically-Centered AI Governance (PCAG)\u003c/b\u003e model and the \u003cb\u003eSustainable Educational Governance Framework (SEG-F)\u003c/b\u003e\u0026mdash;that together conceptualize pedagogy as the ethical core of AI governance.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eBridges abstract ethical principles with actionable governance interventions for educators, policymakers, and designers, contributing to the creation of equitable and sustainable knowledge societies.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eIn doing so, the research advances AI-in-education scholarship from principle-driven description toward mechanism-based understanding, positioning pedagogy as the ethical infrastructure through which AI\u0026rsquo;s societal value in education is realized.\u003c/p\u003e \u003c/div\u003e"},{"header":"2. Theoretical Anchor","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.1. AI as Epistemic Infrastructure: Reshaping the Foundations of Education\u003c/h2\u003e \u003cp\u003eThe concept of knowledge societies (Anheier \u0026amp; Hoelscher, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Stehr, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; UNESCO, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2005\u003c/span\u003e) posits that knowledge is the central resource for sustainable development. In the digital age, Artificial Intelligence (AI) has emerged as a transformative \u003cem\u003eepistemic infrastructure\u003c/em\u003e\u0026mdash;a foundational system that fundamentally alters how knowledge is produced, curated, validated, and disseminated (Floridi, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Williamson, \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Within education, this is not a distant future but a present reality. AI-driven systems\u0026mdash;from intelligent tutoring systems and plagiarism detectors to predictive analytics and automated curriculum generators\u0026mdash;are actively constituting what counts as knowledge, who is considered a knower, and what forms of learning are valued (Knox, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Selwyn, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2021a\u003c/span\u003e). Floridi's (2014) concept of the \"infosphere\" is thus directly operationalized in the classroom and the university, where human-AI interactions are reshaping the educational knowledge ecology. Consequently, for education to remain a cornerstone of a sustainable knowledge society, it must engage with AI not only for its technical efficiency but also, and more importantly, for its profound social, ethical, and epistemic implications, demanding a core focus on inclusiveness and justice (Holmes et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; UNESCO, \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Ethical AI and Inclusivity: From Abstract Principles to Educational Practice\u003c/h2\u003e \u003cp\u003eThe discourse on AI ethics has produced key principles such as beneficence, non-maleficence, autonomy, justice, and explicability (Floridi et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). However, critical scholarship reveals that these principles often remain abstract, failing to address the situated realities of educational practice (Bozkurt, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Crawford, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In an educational context, the principle of \u003cem\u003ejustice\u003c/em\u003e must be translated into a proactive commitment to \u003cem\u003eeducational equity\u003c/em\u003e, ensuring that AI systems do not perpetuate historical biases against marginalized student populations (Baker \u0026amp; Hawn, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). \u003cem\u003eInclusivity\u003c/em\u003e demands more than just access; it requires the active participation of diverse educational stakeholders (teachers, students, parents) in the design and governance of AI systems to avoid structural bias and ensure the benefits of AI are distributed fairly (Jobin et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The OECD (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) rightly emphasizes that AI ethics must be operationalized through concrete governance and regulation, moving beyond mere normative statements to actionable frameworks for schools and universities.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Educational Governance and Pedagogical Value Creation\u003c/h2\u003e \u003cp\u003eFrom an educational leadership perspective, AI governance represents a new frontier for \u003cem\u003eresponsible innovation\u003c/em\u003e (Von Schomberg, \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). It is not merely a defensive risk-management exercise but a strategic opportunity to drive \u003cem\u003epedagogically-sound value creation\u003c/em\u003e. Building on Porter \u0026amp; Kramer's (2011) concept of shared value, ethical and inclusive AI in education expands this horizon by ensuring that technological innovation is intrinsically linked to core educational missions: enhancing human learning, fostering critical thinking, and promoting well-being (Zhang \u0026amp; Aslan, \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). AI governance, therefore, becomes the mechanism to ensure that algorithmic decisions and data-driven practices do not reinforce exclusion or reduce education to metrics, but instead empower educators and promote sustainable, equitable growth for all learners (Williamson et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Educational Equity and Capability Building in the AI Era\u003c/h2\u003e \u003cp\u003eIn education, AI is often framed as a catalyst for personalized learning and universal access (Luckin, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). However, this techno-optimistic view is critically challenged by scholars who warn of \"technological solutionism\"\u0026mdash;the risk of applying AI to complex socio-pedagogical problems without addressing underlying issues of power, context, and inequality (Kritt \u0026amp; Winegar, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Selwyn, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). \u003cem\u003eTrue educational equity in the digital era, as framed by\u003c/em\u003e UNESCO (\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e), \u003cem\u003erequires moving beyond access to technology and towards the development of\u003c/em\u003e capabilities\u003cem\u003e\u0026mdash;the freedom to achieve valued forms of learning and agency\u003c/em\u003e (Sen, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). This entails integrating an equity framework that ensures AI-based education cultivates these capabilities and is aligned with the humanistic goals of SDG 4, rather than undermining them. Through CIMO, ethics is not merely a normative domain but a causal structure that generates sustainability.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e2.5. Integrating CIMO as an Analytical Lens for Educational Inquiry\u003c/h2\u003e \u003cp\u003eTo unify these interconnected theoretical strands into a coherent analytical tool, this study employs the \u003cem\u003eCIMO (Context-Intervention-Mechanism-Outcome)\u003c/em\u003e logic (Denyer et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2008\u003c/span\u003e), specifically tailored for educational research.\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eContext (C): The global educational landscape characterized by datafication, algorithmic bias, digital exclusion, and ethical risks to student autonomy and equity.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eIntervention (I): The application of AI in educational settings, governed by ethical and inclusive principles.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eMechanism (M): The causal forces that explain how the intervention leads to outcomes. This study focuses on educational governance processes, pedagogical value creation, and equity-enhancing capabilities as the core mechanisms.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eOutcome (O): The realization of sustainable, just, and equitable educational environments as the bedrock of sustainable knowledge societies.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eThe CIMO framework facilitates the integration of our multidisciplinary theoretical anchor\u0026mdash;knowledge society theory (Stehr, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; UNESCO, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), AI as epistemic infrastructure (Floridi et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Williamson et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), critical AI ethics (Baker \u0026amp; Hawn, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Bozkurt, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), responsible innovation \u0026amp; value creation (Porter \u0026amp; Kramer, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Von Schomberg, \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), and educational equity as capability building (Luckin, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Selwyn, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2021a\u003c/span\u003e; Sen, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). This synthesis results in a visionary theoretical framework that not only maps the challenges but also provides a causal logic for how ethical and inclusive AI governance can serve as the foundation for sustainable education.\u003c/p\u003e \u003cp\u003eThe integration of these conceptual axes is visualized in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e \u003cem\u003e(see supplementary appendix A)\u003c/em\u003e, which illustrates the dynamic connections between knowledge societies, AI as epistemic infrastructure in education, inclusive ethics, educational governance for pedagogical value, and equity as capability building, all synthesized through the CIMO framework towards the ultimate outcome of sustainable educational ecosystems.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Methodology (Revised Q1-ready version)","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Study Design\u003c/h2\u003e \u003cp\u003eThis study adopts a \u003cem\u003esystematic literature review (SLR)\u003c/em\u003e design to ensure methodological rigor, transparency, and reproducibility in examining the intersection of \u003cem\u003eAI ethics, inclusivity\u003c/em\u003e, and \u003cem\u003esustainability\u003c/em\u003e in education (see supplementary Appendix A Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The review integrates three complementary frameworks:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003ePICo (Population\u0026ndash;Interest\u0026ndash;Context)\u003c/b\u003e to define the conceptual and contextual boundaries of the research;\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003ePRISMA 2020\u003c/b\u003e (Page et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) to structure the literature search, screening, and reporting phases; and\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eCIMO logic\u003c/b\u003e (Denyer et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2008\u003c/span\u003e) to synthesize causal mechanisms linking ethical and inclusive AI interventions to sustainable educational outcomes.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eThe analysis followed a \u003cem\u003ethematic synthesis approach\u003c/em\u003e (Braun \u0026amp; Clarke, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), supported by NVivo 14 software to ensure coding consistency. Methodological rigor was reinforced through intercoder reliability testing (O\u0026rsquo;Connor \u0026amp; Joffe, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), data triangulation, and critical appraisal using established quality assessment tools. This multi-layered approach enables the review not only to aggregate existing evidence but also to uncover the underlying \u003cem\u003emechanisms\u003c/em\u003e through which ethical and inclusive AI can be operationalized across diverse educational contexts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Research Framework: PICo\u003c/h2\u003e \u003cp\u003eThe \u003cb\u003ePICo framework\u003c/b\u003e (Methley et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) was employed to clarify the scope of the research questions and ensure conceptual coherence. Unlike the quantitative-oriented PICO model, PICo is suitable for conceptual, normative, and qualitative inquiries, allowing for exploration of ethical and social dimensions beyond technical performance.\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003ePopulation (P)\u003c/b\u003e: Public, private, and educational organizations implementing AI systems. These entities represent the operational interface where ethical principles are translated into policies and pedagogical practices.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eInterest (I)\u003c/b\u003e: The focus is on ethical principles and inclusivity to identify value systems embedded within AI\u0026rsquo;s design, deployment, and governance processes.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eContext (Co)\u003c/b\u003e: The digital era, characterized by algorithmic decision-making and rapid technological disruption, provides a critical backdrop for examining governance, sustainability, and equity.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eThis framing ensures that the study addresses the \u003cem\u003eAI\u0026ndash;ethics\u0026ndash;inclusivity\u0026ndash;sustainability nexus\u003c/em\u003e, rather than restricting attention to technical or algorithmic dimensions alone.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e3.3. Literature Search Strategy: PRISMA 2020\u003c/h2\u003e \u003cp\u003eFollowing \u003cb\u003ePRISMA 2020\u003c/b\u003e guidelines (Page et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), the literature search prioritized transparency, replicability, and accountability. The search covered seven major databases\u0026mdash;\u003cem\u003eScopus, Web of Science Core Collection, IEEE Xplore, ScienceDirect, SpringerLink, Wiley Online Library\u003c/em\u003e, and \u003cem\u003egrey literature\u003c/em\u003e from international policy organizations (UNESCO, OECD, and the World Bank). This multidisciplinary approach mitigated publication bias and captured diverse epistemic perspectives (Tranfield et al., \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2003\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eA systematic search string was constructed using Boolean and proximity operators aligned with each database\u0026rsquo;s syntax: \u003cem\u003e(\u0026ldquo;artificial intelligence\u0026rdquo; OR \u0026ldquo;AI\u0026rdquo; OR \u0026ldquo;machine learning\u0026rdquo;) AND (\u0026ldquo;ethic*\u0026rdquo; OR \u0026ldquo;responsible AI\u0026rdquo; OR \u0026ldquo;responsible artificial intelligence\u0026rdquo;) AND (\u0026ldquo;inclusiv*\u0026rdquo; OR \u0026ldquo;equit*\u0026rdquo; OR \u0026ldquo;fair*\u0026rdquo; OR \u0026ldquo;accessibility\u0026rdquo;) AND (\u0026ldquo;governance\u0026rdquo; OR \u0026ldquo;policy\u0026rdquo; OR \u0026ldquo;value creation\u0026rdquo; OR \u0026ldquo;education\u0026rdquo; OR \u0026ldquo;sustainable development\u0026rdquo;)\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eThe \u003cem\u003esearch period\u003c/em\u003e was \u003cem\u003eJanuary 2014 to May 2024\u003c/em\u003e, capturing the decade of rapid deep learning advancement and the rise of global ethical-AI debates.\u003c/p\u003e \u003cp\u003eAll identified records were imported into \u003cem\u003eZotero\u003c/em\u003e for deduplication and then screened using \u003cem\u003eRayyan AI\u003c/em\u003e, which enhances efficiency and reduces reviewer bias. Screening was conducted independently by two reviewers, with disagreements resolved through consensus discussion. Intercoder agreement was computed using \u003cb\u003eCohen\u0026rsquo;s κ\u0026thinsp;=\u0026thinsp;0.78\u003c/b\u003e (title/abstract phase) and \u003cb\u003eκ\u0026thinsp;=\u0026thinsp;0.72\u003c/b\u003e (full-text phase), indicating substantial reliability.\u003c/p\u003e \u003cp\u003eThe PRISMA flow (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e in supplementary Appendix A) summarizes the four-stage process:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eIdentification\u003c/b\u003e: 6,785 records retrieved.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eDeduplication\u003c/b\u003e: 1,322 duplicates removed.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eScreening\u003c/b\u003e: 5,463 titles and abstracts assessed; 5,153 excluded as irrelevant.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eEligibility\u003c/b\u003e: 310 full-text articles reviewed; 183 excluded for specific reasons (see below).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eInclusion\u003c/b\u003e: 127 studies retained for final synthesis and thematic analysis.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eReasons for exclusion\u003c/em\u003e included:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;98) \u003cem\u003eIrrelevant focus\u003c/em\u003e \u0026mdash; AI ethics or inclusivity discussed only peripherally.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;45) \u003cem\u003eWrong publication type\u003c/em\u003e \u0026mdash; non-peer-reviewed materials (magazines, theses, or informal reports).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;25) \u003cem\u003eContext mismatch\u003c/em\u003e \u0026mdash; studies unrelated to education or transferable governance contexts.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;15) \u003cem\u003eLack of thematic integration\u003c/em\u003e \u0026mdash; addressed ethics or inclusion but not their interconnection in governance or pedagogy.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eThis systematic reduction\u0026mdash;from 6,785 initial records to 127 eligible studies\u0026mdash;ensures that the synthesis rests exclusively on high-quality, directly relevant literature addressing the intertwined themes of AI ethics, inclusivity, and governance.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e3.4. Data Analysis: Thematic Synthesis\u003c/h2\u003e \u003cp\u003eThe \u003cem\u003ethematic analysis\u003c/em\u003e followed the six-stage procedure by Braun \u0026amp; Clarke (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), enabling systematic pattern identification across heterogeneous studies. NVivo 14 was employed to manage codes, maintain traceability, and visualize emerging themes. The analytical process entailed: (1) Familiarization with data; (2) Generation of initial codes; (3) Searching for themes; (4) Reviewing and refining themes; (5) Defining and naming themes; and (6) Producing the thematic map and synthesis.\u003c/p\u003e \u003cp\u003eReliability was verified through intercoder testing (O\u0026rsquo;Connor \u0026amp; Joffe, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) and peer debriefing among the authors. The thematic synthesis identified four overarching mechanisms that link ethical and inclusive AI practices with sustainable educational outcomes, later elaborated through the CIMO framework.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e3.5. Synthesis Framework: CIMO Logic\u003c/h2\u003e \u003cp\u003eTo move beyond descriptive mapping toward \u003cem\u003eexplanatory synthesis\u003c/em\u003e, thematic results were structured using the \u003cb\u003eCIMO framework\u003c/b\u003e (Denyer et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Rousseau, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2006\u003c/span\u003e):\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eContext (C)\u003c/b\u003e: The global digital ecosystem, characterized by algorithmic bias, inequity, and governance challenges.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eIntervention (I)\u003c/b\u003e: Implementation of AI systems guided by ethical and inclusive principles.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eMechanism (M)\u003c/b\u003e: Governance processes that enable fairness, accountability, transparency, and value creation.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eOutcome (O)\u003c/b\u003e: Sustainable knowledge societies supported by ethical governance and inclusive education.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eThis structure enables identification of \u003cem\u003ecausal mechanisms\u003c/em\u003e linking ethical AI governance to equitable educational outcomes. A \u003cem\u003esensitivity analysis\u003c/em\u003e (Popay et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Tanskanen et al., \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) was conducted to test robustness by:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eRe-running the synthesis after adjusting inclusion/exclusion thresholds; and\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eRe-examining dominant themes after excluding medium-quality studies.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eResults showed minimal variation, confirming analytical stability. Hence, the CIMO logic not only guided synthesis but also enhanced the explanatory validity of findings regarding \u003cem\u003ehow\u003c/em\u003e ethical and inclusive AI contributes to sustainable and just learning ecosystems.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e3.6. Rigor, Validity, and Reproducibility\u003c/h2\u003e \u003cp\u003eTo ensure research integrity, the study adhered to best practices for SLRs in social sciences and educational technology:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eReliability\u003c/b\u003e: Coding reliability established through intercoder agreement (Cohen\u0026rsquo;s κ values reported above).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eValidity\u003c/b\u003e: Triangulation across diverse data sources\u0026mdash;academic articles, organizational reports, and international policy documents.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eCritical Appraisal\u003c/b\u003e: Quality assessed using the \u003cem\u003eJoanna Briggs Institute (JBI)\u003c/em\u003e tool for qualitative/theoretical studies and the \u003cem\u003eMixed Methods Appraisal Tool (MMAT)\u003c/em\u003e for empirical studies (see supplementary Appendix B: Matrix C \u0026amp; C1).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eTransparency\u003c/b\u003e: Every PRISMA 2020 stage (identification, screening, eligibility, inclusion) explicitly documented, including the list of excluded studies and justifications (See Supplentary Appendix A: Matrix C1).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eReproducibility\u003c/b\u003e:All matrix sets (processed data), searches, inclusion/exclusion criteria, coding protocols, and synthesis procedures are archived and made publicly accessible (supplementary Appendix A Matrix A - F) to allow for replication (Tranfield et al., \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2003\u003c/span\u003e).\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eThis rigorous methodological protocol ensures that the review provides both \u003cem\u003eempirical credibility\u003c/em\u003e and \u003cem\u003econceptual depth\u003c/em\u003e, positioning its findings as a reliable foundation for advancing ethical and inclusive AI governance in global education.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Results","content":"\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Characteristics of the Included Studies\u003c/h2\u003e \u003cp\u003eA total of \u003cb\u003e127 studies\u003c/b\u003e met the inclusion criteria (see Supplementary Appendix B, Matrix C2). The corpus demonstrates strong methodological diversity, global reach, and consistent quality, forming a robust evidence base for synthesis.\u003c/p\u003e \u003cp\u003eAs shown in Matrix A (supplementary Appendix B), conceptual and normative works dominate the field (n\u0026thinsp;=\u0026thinsp;38; 30%). These studies\u0026mdash;comprising philosophical essays, normative frameworks, and ethical commentaries\u0026mdash;reflect an evolving scholarly phase in which AI ethics is still establishing its theoretical foundations (Buhmann \u0026amp; Fieseler, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; D. Schiff et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The emphasis remains on defining principles and governance ideals rather than testing empirically grounded hypotheses.\u003c/p\u003e \u003cp\u003eTable A2 shows a strong international distribution: 44 studies (35%) adopt a global or transnational focus, complemented by substantial contributions from the United States (n\u0026thinsp;=\u0026thinsp;16; 13%), the United Kingdom (n\u0026thinsp;=\u0026thinsp;8; 6%), and Australia (n\u0026thinsp;=\u0026thinsp;8; 6%). Importantly, representation from the Global South is emerging through studies from Latin America (Gutierrez et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Gutierrez Y Restrepo \u0026amp; Floris, 2022), Brazil (Cantarini, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), India (Sharma et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and Africa (Spiegel et al., \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), signaling a gradual diversification of epistemic voices beyond traditional Western-centric narratives.\u003c/p\u003e \u003cp\u003eAs reported in Matric C \u0026amp; C1, quality assessment used the \u003cb\u003eJBI checklists\u003c/b\u003e (n\u0026thinsp;=\u0026thinsp;115; 82.1%) for normative and theoretical works and the \u003cb\u003eMMAT\u003c/b\u003e (n\u0026thinsp;=\u0026thinsp;25; 17.9%) for empirical studies. This dual-tool strategy ensured comparability across methodological traditions and maintained evaluative rigor. The overall methodological quality is high: 72.1% of studies scored\u0026thinsp;\u0026ge;\u0026thinsp;70 on a 100-point scale (average\u0026thinsp;\u0026asymp;\u0026thinsp;73.5%). No study was excluded for low quality, though 23 (18.1%) scored below 65%.\u003c/p\u003e \u003cp\u003eMattrix D indicates strong thematic interconnectedness. Few papers address a single research question (RQ1 only: n\u0026thinsp;=\u0026thinsp;9; RQ2 only: n\u0026thinsp;=\u0026thinsp;7; RQ3 only: n\u0026thinsp;=\u0026thinsp;3). Most contribute to multiple questions\u0026mdash;particularly RQ2 \u0026amp; RQ3 (n\u0026thinsp;=\u0026thinsp;41; 29.3%) and RQ1 \u0026amp; RQ3 (n\u0026thinsp;=\u0026thinsp;37; 26.4%)\u0026mdash;underscoring the multidisciplinary character of the field and the entanglement between ethical theory, governance mechanisms, and educational practice.\u003c/p\u003e \u003cp\u003eOverall, these characteristics confirm a \u003cem\u003emethodologically sound and globally distributed evidence base\u003c/em\u003e, enabling reliable synthesis of mechanisms linking AI ethics, inclusivity, and sustainability in education.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Framing the Relationship between AI Ethics and Inclusivity (RQ1)\u003c/h2\u003e \u003cp\u003eAcross the international literature, AI ethics and inclusivity are framed not as linear cause-effect phenomena but as interdependent socio-technical constructs underpinning \u003cem\u003esustainable knowledge societies\u003c/em\u003e. Four dominant framings emerge \u003cb\u003e(\u003c/b\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e2\u003c/span\u003e\u003cb\u003e)\u003c/b\u003e, each addressing a distinct dimension of the ethical\u0026ndash;inclusive AI nexus.\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eCritical\u0026ndash;Decolonial Framing\u003c/b\u003e: Studies such as Cantarini (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), Casillo et al. (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), Sijing \u0026amp; Lan (\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), and Xu (\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) highlight epistemic injustice and data colonialism as structural barriers to inclusivity. Here, AI ethics is conceptualized as a \u003cem\u003edecolonial project\u003c/em\u003e requiring contextualized, locally grounded, and sovereignty-respecting AI systems.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eLegal\u0026ndash;Institutional Framing\u003c/b\u003e: Ess et al. (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), Price \u0026amp; Cohen (\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), Smith et al. (2020), and Su (\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) view inclusivity as an outcome of enforceable legal and institutional safeguards\u0026mdash;such as the \u003cem\u003eright to explanation\u003c/em\u003e and data-protection mechanisms (e.g., GDPR)\u0026mdash;that embed accountability and human oversight into AI governance.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eCapability\u0026ndash;Pedagogical Framing (The Educational Core)\u003c/b\u003e: Educational research (Bozkurt, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Calo, 2018; Corredor Garc\u0026iacute;a et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Rocco et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) frames AI ethics as a \u003cem\u003elearnable capability\u003c/em\u003e cultivated through curriculum integration, teacher training, and critical AI literacy. Inclusivity thus becomes a pedagogical outcome rather than a compliance requirement, establishing ethics as a core educational competence.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eValue-Based Governance Framing\u003c/b\u003e: Gorur et al. (2020), Keeble \u0026amp; Blatchly-Lewis (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and Tsafack Chetsa (2021) link ethics and inclusivity to shared-value creation and ecological sustainability, arguing that ethical governance generates inclusivity, which in turn reinforces social legitimacy and long-term system viability.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eSynthesis\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eTogether, these framings demonstrate a shift from moral abstraction toward practical operationalization. The pedagogical-capability framing emerges as the \u003cem\u003etranslational core\u003c/em\u003e\u0026mdash;the pathway through which ethical principles become lived educational realities. Other framings provide the structural and critical scaffolding that enables pedagogy to function as the ethical engine of sustainable learning.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eFramings of Ethical and Inclusive AI\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFraming\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDefinitional Focus\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003en (studies)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eKey References\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCritical\u0026ndash;Decolonial / Structural\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eChallenges epistemic injustice and data colonialism; emphasizes local context, plural epistemologies, and data sovereignty.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCantarini (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); Casillo et al. (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e); Erman \u0026amp; Furendal (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLegal\u0026ndash;Institutional\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLinks AI ethics with law, regulation, and accountability frameworks ensuring transparency and human oversight.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEss et al. (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2018\u003c/span\u003e); Price \u0026amp; Cohen (\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2019\u003c/span\u003e); Su (\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCapability\u0026ndash;Pedagogical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFrames ethics as social capacity; focuses on education, AI literacy, and empowerment for inclusive participation.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCorredor Garc\u0026iacute;a et al. (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e); Rocco et al. (\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); Bogina et al. (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eValue-Based Sustainability\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eConnects ethical AI with shared value creation and ecological responsibility in organizational contexts.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eKeeble \u0026amp; Blatchly-Lewis (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); Sharma et al. (\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Mechanisms Linking AI to Value Creation, Governance, and Educational Equity (RQ2)\u003c/h2\u003e \u003cp\u003eApplying the \u003cb\u003eCIMO\u003c/b\u003e synthesis revealed four classes of mechanisms through which ethical and inclusive AI contributes to equitable and sustainable education. Their relative importance varies by context, but in educational environments the \u003cem\u003ePedagogical-Capacity Mechanism\u003c/em\u003e is consistently dominant.\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003ePedagogical-Capacity Mechanisms\u003c/b\u003e: Interventions such as ethics-infused curricula (Corredor Garc\u0026iacute;a et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Rocco et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), teacher professional development (Calo, 2018), and AI literacy programs (Selwyn, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2021a\u003c/span\u003e; Walsh et al., 2022) increase ethical awareness and behavioral competence. These mechanisms cultivate the human capabilities essential for equitable participation in digital societies.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eTechnical-Process Controls\u003c/b\u003e: Mechanisms including fairness metrics (Paic, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Weber, 2019), explainable-AI techniques (Castelli et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and audit trails (Weber-Lewerenz, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) enhance transparency and reduce algorithmic bias. Such controls build trust in educational technologies, enabling responsible adoption across sectors (Ahmad et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Lee, 2020).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eRegulatory-Accountability Mechanisms\u003c/b\u003e: Legal standards and compliance instruments (Kim et al., 2021; Sapienza, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Smith et al., 2020) enforce transparency and rights protection, creating a governance environment conducive to long-term value creation and safeguarding ethical experimentation in education.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eParticipatory-Legitimation Mechanisms\u003c/b\u003e: Inclusive design and deliberative approaches (Camar\u0026eacute;na, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Bhawra et al., 2021; K\u0026auml;stner \u0026amp; Kang, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) integrate community knowledge, enhance contextual relevance, and strengthen social legitimacy (Gutierrez Y Restrepo \u0026amp; Floris, 2022).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eSynthesis\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eEthical AI\u0026rsquo;s contribution to value and equity is \u003cem\u003emediated\u003c/em\u003e rather than automatic. Pedagogical capacity functions as the primary driver, while technical, regulatory, and participatory mechanisms provide the enabling infrastructure\u0026mdash;trust, accountability, and legitimacy\u0026mdash;that make transformative pedagogy possible.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMechanisms for Operationalizing Ethical and Inclusive AI\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMechanism\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFunctional Description\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003en (studies)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eKey References\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTechnical\u0026ndash;Process Controls\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEmbeds fairness metrics, explainability (XAI), bias audits, and lifecycle data governance into design.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePaic (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2019\u003c/span\u003e); Castelli et al. (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRegulatory\u0026ndash;Accountability\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEstablishes standards, certification, and liability frameworks to enforce transparency and compliance.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePrice \u0026amp; Cohen (\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2019\u003c/span\u003e); Sapienza (\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); Losavio (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2021\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eParticipatory\u0026ndash;Legitimation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUses co-design and multi-stakeholder dialogue to ensure inclusion and social legitimacy.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCamar\u0026eacute;na (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2021\u003c/span\u003e); K\u0026auml;stner \u0026amp; Kang (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e); Dennehy et al. (2021)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePedagogical\u0026ndash;Capacity Building\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIntegrates ethics into education, training, and digital literacy to strengthen long-term inclusion.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCorredor Garc\u0026iacute;a et al. (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e); Rocco et al. (\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); Khoury et al. (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eSummary\u003c/strong\u003e \u003cp\u003eThese mechanisms operate symbiotically\u0026mdash;technical and regulatory actions provide structural reliability, while participatory and pedagogical initiatives ensure inclusivity and accountability. Together, they form the \u003cem\u003eoperational backbone\u003c/em\u003e of ethical AI governance.\u003c/p\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003e4.4. Opportunities, Challenges, and Pathways toward Global Governance (RQ3)\u003c/h2\u003e \u003cp\u003e \u003cem\u003eEmerging Opportunities.\u003c/em\u003e \u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eConvergence of principles\u003c/em\u003e\u0026mdash;a global normative consensus on fairness, transparency, and non-maleficence (Kim et al., 2021; Kiritchenko et al., 2021; Tsafack Chetsa, 2021).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eSectoral policy innovation\u003c/em\u003e\u0026mdash;context-specific ethical guidelines in domains such as healthcare (Khoury et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), automotive (Luetge et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), and smart cities (Keeble \u0026amp; Blatchly-Lewis, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) offer scalable blueprints.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eEthical-enabler technologies\u003c/em\u003e\u0026mdash;progress in explainable AI, fairness auditing, and differential privacy operationalizes ethics in practice.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eGlobal-South capacity building\u003c/em\u003e\u0026mdash;contextual models from Latin America, India, and Africa (Cantarini, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Gutierrez Y Restrepo \u0026amp; Floris, 2022; Sijing \u0026amp; Lan, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) democratize governance and rebalance epistemic authority.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003ePersistent Challenges.\u003c/em\u003e \u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003ePrinciple\u0026ndash;practice gap\u003c/em\u003e\u0026mdash;ethical declarations often lack operational translation, resulting in \u0026ldquo;ethics washing\u0026rdquo; (D. Schiff et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eRegulatory fragmentation\u003c/em\u003e\u0026mdash;divergent jurisdictional frameworks create accountability gaps and compliance complexity (Sapienza, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Su, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003ePower asymmetry\u003c/em\u003e\u0026mdash;policy-making remains dominated by Global-North actors and industry (D. S. Schiff et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eInfrastructural bias\u003c/em\u003e\u0026mdash;historical data imbalances perpetuate inequity (Denton et al., 2021; Quintarelli et al., 2019).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eToward a Hybrid Global Governance Framework\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eThe literature collectively advocates a multi-level architecture:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eGlobal Normative Layer\u003c/em\u003e: universal human-rights-based principles guiding all AI use.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eSectoral Operational Layer\u003c/em\u003e: domain-specific standards and certification systems (e.g., education, health).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eLocal Implementation Layer\u003c/em\u003e: participatory and context-adaptive mechanisms ensuring legitimacy.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eSupporting Mechanisms\u003c/em\u003e: investment in Global-South capacity, open data, and independent audits.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eSynthesis.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eEffective ethical-AI governance requires \u003cem\u003evertical coherence\u003c/em\u003e\u0026mdash;aligning top-down principles with bottom-up contextualization. Evidence across 69 studies shows that enduring ethical performance depends on integration among these layers (Floridi \u0026amp; Cowls, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; McKay et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eGovernance Themes Emerging from the Literature\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGovernance Theme\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDefinitional Focus\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003en (studies)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eKey References\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHybrid Global Framework\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAligns universal ethical principles with sectoral modules and local feedback loops.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFloridi \u0026amp; Cowls (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2021\u003c/span\u003e); Pitt et al. (2019)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSectoral Operationalization\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTranslates principles into sector-specific standards (e.g., education, health, smart cities).\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMcKay et al. (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); Luetge et al. (\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2021\u003c/span\u003e); Keeble \u0026amp; Blatchly-Lewis (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLocal / Decolonial Adaptation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEmbeds governance within local epistemic contexts to protect data sovereignty and justice.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCantarini (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); Gutierrez Y Restrepo \u0026amp; Floris (2022)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStandards, Audits \u0026amp; Oversight\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePromotes certification, interoperability, and independent auditing to reinforce trust.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFrischknecht-Gruber et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); Schiff et al. (\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e\u003cb\u003eSummary\u003c/b\u003e\u003c/p\u003e\u003cp\u003eEffective ethical AI governance requires \u003cem\u003emulti-level coherence\u003c/em\u003e\u0026mdash;balancing top-down standardization with bottom-up contextualization. Alignment across these levels mitigates fragmentation, reinforces accountability, and sustains equity within global AI ecosystems.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003e4.5. Integrative Synthesis: Themes, Mechanisms, and Strategic Relevance\u003c/h2\u003e \u003cp\u003eThe cross-theme synthesis demonstrates that \u003cem\u003eethical AI governance is intrinsic to organizational and educational strategy\u003c/em\u003e, not an external compliance activity. Four mechanisms\u0026mdash;\u003cem\u003eTransparency\u003c/em\u003e \u0026rarr; \u003cem\u003eTrust, Accountability\u003c/em\u003e \u0026rarr; \u003cem\u003eAlignment, Inclusivity\u003c/em\u003e \u0026rarr; \u003cem\u003eCapability, and Sustainability\u003c/em\u003e \u0026rarr; \u003cem\u003eValue Creation\u003c/em\u003e\u0026mdash;constitute a cyclical system of strategic knowledge governance:\u003c/p\u003e \u003cp\u003e \u003cb\u003eData \u0026rarr; Governance \u0026rarr; Innovation \u0026rarr; Sustainability \u0026rarr; Data\u003c/b\u003e \u003c/p\u003e \u003cp\u003eWithin this loop, ethical intelligence functions as a dynamic capability driving continuous learning, stakeholder trust, and adaptive performance. Ethical AI thus becomes a \u003cem\u003estrategic infrastructure\u003c/em\u003e for sustainable knowledge societies.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSynthesis of Thematic Findings, Mechanisms, and Managerial Implications\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTheme\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCore Mechanism (CIMO Synthesis)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eEmpirical Focus / Illustrative Examples\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eManagerial and Strategic Relevance\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003e1. Ethical Foundations and Principles\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTransparency \u0026rarr; Trust\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eExplainable AI tools, open audit trails, and fairness metrics increase institutional legitimacy (Chakraborty et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Helberger et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBuilds stakeholder confidence, enhances data integrity, and establishes the ethical baseline for responsible innovation.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003e2. Organizational and Governance Mechanisms\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAccountability \u0026rarr; Alignment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMulti-level governance structures, algorithmic audits, and Corporate Digital Responsibility frameworks (Shneiderman, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Weber-Lewerenz, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEmbeds ethical oversight into strategic and operational processes, aligning AI objectives with mission and regulatory standards.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003e3. Strategic and Operational Integration\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eInclusivity \u0026rarr; Capability\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eParticipatory innovation design, ethics-by-design processes, and inclusive data governance (Buhmann \u0026amp; Fieseler, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Rajagopal et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTransforms ethics into organizational capability, fostering innovation readiness, diversity, and resilience.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003e4. Global and Cross-Sectoral Governance Models\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSustainability \u0026rarr; Value Creation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eUNESCO, OECD, and sectoral frameworks in healthcare, education, and agriculture (Ahmad et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Klerkx \u0026amp; Rose, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLinks ethical governance to measurable sustainability outcomes, translating social responsibility into long-term competitive advantage.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section2\"\u003e \u003ch2\u003e4.6. Cross-RQ Integration and Conceptual Continuity\u003c/h2\u003e \u003cp\u003eThe findings across RQ1\u0026ndash;RQ3 form a coherent continuum:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eFramings of ethics and inclusivity (RQ1)\u003c/em\u003e provide the \u003cem\u003enormative foundation\u003c/em\u003e;\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eMechanisms of governance and value creation (RQ2)\u003c/em\u003e constitute the \u003cem\u003eoperational process\u003c/em\u003e; and\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cem\u003eOpportunities and global governance pathways (RQ3)\u003c/em\u003e delineate the \u003cem\u003estrategic horizon\u003c/em\u003e.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eThis interconnection validates the study\u0026rsquo;s core argument: achieving sustainable knowledge societies requires the interplay of ethical framings, operational mechanisms, and multi-level governance coherence. The synthesis directly informs the construction of the \u003cb\u003ePedagogically-Centered AI Governance (PCAG)\u003c/b\u003e and \u003cb\u003eSustainable Educational Governance Framework (SEG-F)\u003c/b\u003e models \u003cb\u003e(see\u003c/b\u003e Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003e \u003cb\u003ein supplemtary A)\u003c/b\u003e, which integrate theoretical, empirical, and normative dimensions into a replicable architecture for global education systems.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Discussion and Conceptual Synthesis","content":"\u003ch2\u003e5.1. Reframing AI Ethics as Pedagogical Infrastructure\u003c/h2\u003e\n\u003cp\u003eThe findings collectively reframe Artificial Intelligence not merely as a technological add-on but as a\u0026nbsp;\u003cstrong\u003e\u003cem\u003epedagogical infrastructure\u003c/em\u003e\u003c/strong\u003e\u0026mdash;a foundation that shapes how knowledge is produced, distributed, and governed (Floridi, 2014; Selwyn, 2021b; Holmes et al., 2022).\u003c/p\u003e\n\u003cp\u003eThe global literature reveals a transition from normative reflection toward \u003cem\u003efunctional operationalization\u003c/em\u003e: AI ethics now operates as a \u003cstrong\u003e\u003cem\u003econstitutive condition for equitable learning ecosystems\u003c/em\u003e\u003c/strong\u003e, rather than as an external constraint (Buhmann \u0026amp; Fieseler, 2021; Eynon, 2022).\u003c/p\u003e\n\u003cp\u003eThis reframing aligns with the \u003cem\u003ecritical\u0026ndash;decolonial\u003c/em\u003e and \u003cem\u003ecapability\u0026ndash;pedagogical\u003c/em\u003e framings (RQ1), where inclusivity and fairness are not end goals but\u0026nbsp;\u003cstrong\u003e\u003cem\u003epedagogical capabilities\u003c/em\u003e\u003c/strong\u003e to be cultivated (Bozkurt, 2023; Corredor Garc\u0026iacute;a et al., 2021).\u003c/p\u003e\n\u003cp\u003eHence, ethics and inclusivity become dynamic processes embedded in \u003cstrong\u003e\u003cem\u003ecurriculum design, teacher education, and AI literacy\u003c/em\u003e\u003c/strong\u003e, directly influencing learners\u0026rsquo; epistemic agency and social participation.\u003c/p\u003e\n\u003cp\u003eIn this sense, AI functions as a \u003cstrong\u003e\u003cem\u003emeta-learning system\u003c/em\u003e\u003c/strong\u003e\u0026mdash;an epistemic infrastructure that governs \u003cem\u003ehow\u003c/em\u003e educational value is created and \u003cem\u003ewho\u003c/em\u003e participates in it (Williamson \u0026amp; Eynon, 2020; Zawacki-Richter et al., 2019). This reconceptualization grounds the \u003cstrong\u003ePedagogically-Centered AI Governance (PCAG)\u003c/strong\u003e model proposed in this study.\u003c/p\u003e\n\u003ch2\u003e5.2. The Pedagogically-Centered AI Governance (PCAG) Model\u003c/h2\u003e\n\u003cp\u003eThe \u003cstrong\u003ePCAG model\u003c/strong\u003e synthesizes results from RQ1 and RQ2 into a multi-mechanistic framework that explains\u0026nbsp;\u003cem\u003ehow ethical and inclusive AI generates educational and societal value\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003eIt situates \u003cstrong\u003e\u003cem\u003epedagogical capacity-building\u003c/em\u003e\u003c/strong\u003e as the central mechanism (CIMO synthesis), supported by three enabling infrastructures: \u003cem\u003etechnical-process controls\u003c/em\u003e, \u003cem\u003eregulatory-accountability systems\u003c/em\u003e, and \u003cem\u003eparticipatory-legitimation structures\u003c/em\u003e (see Table 3).\u003c/p\u003e\n\u003cp\u003e(1) Pedagogical Core\u003c/p\u003e\n\u003cp\u003eAt its heart, PCAG emphasizes\u0026nbsp;\u003cem\u003eAI ethics as a learnable and teachable capability\u003c/em\u003e. When ethics is integrated into the curriculum, professional training, and digital literacy programs, it fosters collective epistemic resilience (Rocco et al., 2022; Calo, 2018; Walsh et al., 2022).\u003c/p\u003e\n\u003cp\u003eThis \u0026ldquo;pedagogical core\u0026rdquo; transforms normative principles\u0026mdash;fairness, transparency, accountability\u0026mdash;into \u003cem\u003esocial capabilities\u003c/em\u003e that underpin sustainable learning societies (Bozkurt, 2023; Gutierrez Y Restrepo \u0026amp; Floris, 2022).\u003c/p\u003e\n\u003cp\u003e(2) Technical and Regulatory Infrastructure\u003c/p\u003e\n\u003cp\u003eTechnical and regulatory mechanisms establish the\u0026nbsp;\u003cstrong\u003e\u003cem\u003etrust architecture\u003c/em\u003e\u003c/strong\u003e necessary for ethical learning systems. Fairness metrics (Paic, 2019), explainable AI (Castelli et al., 2024), and algorithmic audits (Weber-Lewerenz, 2021) operationalize transparency, while legal frameworks (Smith et al., 2020; Sapienza, 2022) institutionalize accountability.\u003c/p\u003e\n\u003cp\u003eTogether, they create a stable environment where educational actors can engage safely with AI, bridging the \u003cem\u003eprinciple\u0026ndash;practice gap\u003c/em\u003e that often undermines ethical implementation (D. Schiff et al., 2021).\u003c/p\u003e\n\u003cp\u003e(3) Participatory and Decolonial Anchors\u003c/p\u003e\n\u003cp\u003eThe participatory mechanisms identified (Camar\u0026eacute;na, 2021; Bhawra et al., 2021; K\u0026auml;stner \u0026amp; Kang, 2020) ensure that AI development in education remains\u0026nbsp;\u003cstrong\u003e\u003cem\u003elocally relevant, culturally responsive, and epistemically just\u003c/em\u003e\u003c/strong\u003e (Cantarini, 2022; Casillo et al., 2024).\u003c/p\u003e\n\u003cp\u003eBy embedding co-design and local deliberation, PCAG counteracts the \u003cem\u003eGlobal North bias\u003c/em\u003e in AI governance (D. S. Schiff et al., 2022) and affirms the agency of educators and learners as co-creators of ethical technology.\u003c/p\u003e\n\u003cp\u003e(4) Systemic Integration\u003c/p\u003e\n\u003cp\u003eThese three subsystems\u0026mdash;pedagogical, technical\u0026ndash;regulatory, and participatory\u0026mdash;form an interdependent architecture.\u003c/p\u003e\n\u003cp\u003ePedagogical capacity provides the \u003cem\u003eagency layer\u003c/em\u003e, technical and regulatory structures supply the \u003cem\u003eintegrity layer\u003c/em\u003e, and participatory mechanisms contribute the\u0026nbsp;\u003cem\u003elegitimacy layer\u003c/em\u003e.\u003cbr\u003eTheir interaction constitutes the \u003cstrong\u003e\u003cem\u003eethical\u0026ndash;inclusive AI ecosystem\u003c/em\u003e\u003c/strong\u003e capable of generating trust, equity, and sustainable innovation (Ahmad et al., 2021; Tsafack Chetsa, 2021).\u003c/p\u003e\n\u003cp\u003eFigure 4 visualizes this integrative model: ethics operates through pedagogy, is stabilized by governance, and validated through participation.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e5.3. Integrated Framework: From Pedagogy to Global Governance\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe integration of the \u003cstrong\u003ePedagogically-Centered AI Governance (PCAG)\u003c/strong\u003e and the \u003cstrong\u003eSustainable Educational Governance Framework (SEG-F)\u003c/strong\u003e represents a holistic architecture that connects micro-level educational mechanisms with macro-level global governance systems. This integration explains \u003cem\u003ehow ethical and inclusive AI transitions from classroom pedagogy to global sustainability structures.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAt the \u003cstrong\u003e\u003cem\u003epedagogical level\u003c/em\u003e\u003c/strong\u003e, PCAG situates ethics as a \u003cem\u003elearnable capability\u003c/em\u003e embedded within educational design, teacher training, and AI literacy programs (Bozkurt, 2023; Corredor Garc\u0026iacute;a et al., 2021; Rocco et al., 2022). Here, ethics functions not as a regulatory constraint but as a transformative learning process that empowers learners and educators to critically engage with technology. This process forms the \u003cem\u003ecapability-building foundation\u003c/em\u003e for sustainable digital societies (Calo, 2018; Walsh et al., 2022).\u003c/p\u003e\n\u003cp\u003eAt the \u003cstrong\u003e\u003cem\u003einstitutional and regulatory level\u003c/em\u003e\u003c/strong\u003e, the SEG-F framework extends these pedagogical foundations into broader governance coherence. It aligns \u003cem\u003euniversal ethical principles\u003c/em\u003e (fairness, accountability, transparency) with sectoral operationalization and local adaptation (Floridi \u0026amp; Cowls, 2021; McKay et al., 2022; UNESCO, 2021). The framework ensures that ethical AI practices in education are supported by technical-process controls (Paic, 2019; Castelli et al., 2024), enforceable standards (Smith et al., 2020; Sapienza, 2022), and participatory feedback loops that sustain contextual legitimacy (Camar\u0026eacute;na, 2021; Gutierrez Y Restrepo \u0026amp; Floris, 2022).\u003c/p\u003e\n\u003cp\u003eTogether, these two frameworks form a \u003cstrong\u003e\u003cem\u003econtinuum of governance learning systems\u003c/em\u003e\u003c/strong\u003e. PCAG functions as the \u003cem\u003epedagogical engine\u003c/em\u003e that internalizes ethics and inclusivity into learning, while SEG-F provides the \u003cem\u003estructural ecosystem\u003c/em\u003e that scales these values through multi-level governance. The integration operationalizes \u003cstrong\u003e\u003cem\u003eFloridi\u0026rsquo;s concept of the infosphere\u003c/em\u003e\u003c/strong\u003e\u0026mdash;where knowledge, data, and ethics co-evolve within an interconnected digital ecology\u0026mdash;and extends \u003cstrong\u003e\u003cem\u003eSen\u0026rsquo;s capability approach\u003c/em\u003e\u003c/strong\u003e by demonstrating how ethical governance expands individuals\u0026rsquo; real freedoms to learn, participate, and innovate within AI-mediated environments.\u003c/p\u003e\n\u003cp\u003eThis convergence thus reframes AI ethics as both \u003cstrong\u003e\u003cem\u003eepistemic and systemic infrastructure\u003c/em\u003e\u003c/strong\u003e. Ethical AI in education is no longer confined to compliance documents or curricular modules; it becomes a \u003cem\u003edynamic architecture\u003c/em\u003e that links pedagogy, governance, and sustainability. By embedding human capability and moral intelligence within the digital ecosystem, the integrated PCAG\u0026ndash;SEG-F framework provides a blueprint for building \u003cstrong\u003e\u003cem\u003eethical knowledge societies\u003c/em\u003e\u003c/strong\u003e that are simultaneously locally grounded and globally coherent.\u003c/p\u003e\n\u003ch2\u003e5.4. Policy and Practice Implications\u003c/h2\u003e\n\u003cp\u003eThe synthesis has direct implications for educational policymakers, institutional leaders, and AI developers.\u003c/p\u003e\n\u003col\u003e\n \u003cli\u003e\u003cstrong\u003ePolicy Integration:\u0026nbsp;\u003c/strong\u003eEthical AI must be embedded in\u0026nbsp;\u003cstrong\u003e\u003cem\u003eeducation policy and accreditation standards\u003c/em\u003e\u003c/strong\u003e, linking digital transformation with social inclusion (UNESCO, 2023).\u003cbr\u003e\u0026nbsp;Ministries of education should integrate AI literacy and data ethics into teacher competency frameworks (Walsh et al., 2022; Corredor Garc\u0026iacute;a et al., 2021).\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eInstitutional Governance:\u0026nbsp;\u003c/strong\u003eUniversities and schools should adopt \u003cstrong\u003e\u003cem\u003emulti-level governance structures\u003c/em\u003e\u003c/strong\u003e\u0026mdash;combining AI ethics committees, transparent data policies, and participatory audit mechanisms\u0026mdash;to ensure accountability and contextual responsiveness (Weber-Lewerenz, 2021; Shneiderman, 2020).\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eTechnological Co-Design:\u0026nbsp;\u003c/strong\u003eDevelopers should move from user-centered to \u003cem\u003eeducator-centered\u003c/em\u003e and \u003cem\u003ecommunity-centered\u003c/em\u003e design, engaging teachers and learners in co-creation processes that embed inclusivity from inception (Camar\u0026eacute;na, 2021; Bhawra et al., 2021).\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eGlobal Collaboration:\u0026nbsp;\u003c/strong\u003eInternational organizations (OECD, UNESCO, EU) should facilitate \u003cstrong\u003e\u003cem\u003eSouth\u0026ndash;South knowledge exchange\u003c/em\u003e\u003c/strong\u003e, supporting capacity building and policy harmonization to address epistemic inequality in AI governance (Cantarini, 2022; Gutierrez et al., 2022).\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eCollectively, these implications advance a \u003cem\u003epractical roadmap\u003c/em\u003e for operationalizing ethical AI governance as an educational and societal infrastructure.\u003c/p\u003e\n\u003ch2\u003e5.5. Conceptual Synthesis and Future Directions\u003c/h2\u003e\n\u003cp\u003eThe synthesis confirms that \u003cstrong\u003e\u003cem\u003eethics and inclusivity are not adjuncts but core architectures\u003c/em\u003e\u003c/strong\u003e of sustainable AI-driven education.\u003c/p\u003e\n\u003cp\u003ePCAG provides the micro-foundations\u0026mdash;pedagogy, literacy, governance\u0026mdash;while SEG-F establishes the macro-structure\u0026mdash;principles, standards, and global coherence.\u003c/p\u003e\n\u003cp\u003eFuture research should empirically test these models across diverse educational systems, focusing on:\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003elongitudinal studies of AI ethics curricula and learning outcomes,\u003c/li\u003e\n \u003cli\u003ecomparative analyses of governance structures across regions, and\u003c/li\u003e\n \u003cli\u003eparticipatory design methodologies that integrate Global South epistemologies.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eThrough such expansion, ethical and inclusive AI can evolve from discourse to \u003cstrong\u003e\u003cem\u003einfrastructure\u003c/em\u003e\u003c/strong\u003e, ensuring that the next phase of digital transformation is not only intelligent but also just, human-centered, and pedagogically grounded.\u003c/p\u003e"},{"header":"6. Conclusion and Recommendations","content":"\u003cp\u003eArtificial Intelligence has become a constitutive element of educational transformation\u0026mdash;no longer a peripheral tool, but a \u003cem\u003egoverning infrastructure\u003c/em\u003e that shapes how knowledge, capability, and equity are produced. This review demonstrates that the ethical and inclusive governance of AI is not a secondary concern, but the \u003cem\u003eepistemic foundation\u003c/em\u003e of sustainable knowledge societies. Synthesizing evidence from 127 peer-reviewed studies through PRISMA, PICo, and CIMO frameworks, the research conceptualized two interlocking models: The \u003cb\u003ePedagogically-Centered AI Governance (PCAG)\u003c/b\u003e and the \u003cb\u003eSustainable Educational Governance Framework (SEG-F)\u003c/b\u003e. Together, they explain \u003cem\u003ehow\u003c/em\u003e and \u003cem\u003ewhy\u003c/em\u003e ethical principles translate into educational value creation\u0026mdash;by aligning pedagogical capacity, regulatory integrity, and participatory legitimacy.\u003c/p\u003e \u003cp\u003eThe findings advance three critical insights. First, \u003cem\u003epedagogy is the operational core\u003c/em\u003e of AI ethics: inclusive and critical AI literacy must be cultivated through curriculum, teacher education, and lifelong learning systems. Second, \u003cem\u003etechnical and regulatory infrastructures\u003c/em\u003e\u0026mdash;fairness metrics, algorithmic audits, and accountability frameworks\u0026mdash;function as enablers that stabilize ethical practice across institutions. Third, \u003cem\u003eparticipatory and decolonial approaches\u003c/em\u003e ensure that AI systems reflect diverse epistemologies, countering the dominance of Global North narratives in AI governance. These interdependent mechanisms redefine AI ethics from a declarative code into a \u003cem\u003epedagogical process\u003c/em\u003e of empowerment, deliberation, and value creation.\u003c/p\u003e \u003cp\u003eFrom a policy perspective, the study recommends embedding AI ethics into national education strategies and accreditation standards, supported by capacity-building programs and independent audit mechanisms. Global organizations such as UNESCO, OECD, and regional alliances should operationalize the SEG-F by promoting \u003cem\u003emulti-level coherence\u003c/em\u003e\u0026mdash;linking global normative principles with sectoral adaptation and local implementation. Equally, institutions should establish internal governance architectures that integrate AI ethics committees, open-data charters, and participatory design platforms to ensure transparency, inclusivity, and accountability. Through these mechanisms, ethical AI can evolve from regulatory compliance into a \u003cem\u003estrategic infrastructure for sustainable education\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eTheoretically, this review contributes to the growing discourse on epistemic sustainability\u0026mdash;showing that just as infrastructures determine access to energy or transport, \u003cem\u003eAI infrastructures determine access to knowledge and opportunity\u003c/em\u003e. Future research should empirically test the PCAG and SEG-F models across diverse contexts, using longitudinal and comparative methodologies that examine their impact on learning outcomes, institutional governance, and social equity. Methodological innovations\u0026mdash;such as mixed-methods AI audits, participatory action research, and design-based policy experiments\u0026mdash;will be critical to deepen causal understanding. Ultimately, the ethical future of education depends not on how intelligent our machines become, but on how wisely and inclusively we design the human systems that govern them. Positioning AI governance as a pedagogical process redefines education as the moral infrastructure of the digital age.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eThe author confirms sole responsibility for the following: study conception and design, data collection, analysis and interpretation of results, and manuscript preparation. The author reviewed and approved the final version of the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eData Availability StatementAll data supporting the conclusions in this article are presented within the article itself and in the accompanying Supplementary Appendix. This appendix includes the search matrix, inclusion/exclusion criteria, and a complete data synthesis table.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAhmad, M. A., Overman, S., Allen, C., Kumar, V., Teredesai, A., \u0026amp; Eckert, C. (2021). \u003cem\u003eSoftware as a Medical Device: Regulating AI in Healthcare via Responsible AI\u003c/em\u003e. 4023\u0026ndash;4024. https://doi.org/10.1145/3447548.3470823\u003c/li\u003e\n\u003cli\u003eAnheier, H. K., \u0026amp; Hoelscher, M. (2015). The 2005 UNESCO Convention and Civil Society: An Initial Assessment. In \u003cem\u003eGlobalization, Culture, and Development\u003c/em\u003e (pp. 182\u0026ndash;202). Palgrave Macmillan UK. https://doi.org/10.1057/9781137397638_13\u003c/li\u003e\n\u003cli\u003eBaker, R. S. (2016). Stupid Tutoring Systems, Intelligent Humans. \u003cem\u003eInternational Journal of Artificial Intelligence in Education\u003c/em\u003e, \u003cem\u003e26\u003c/em\u003e(2), 600\u0026ndash;614. https://doi.org/10.1007/s40593-016-0105-0\u003c/li\u003e\n\u003cli\u003eBaker, R. S., \u0026amp; Hawn, A. (2022). Algorithmic Bias in Education. \u003cem\u003eInternational Journal of Artificial Intelligence in Education\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(4), 1052\u0026ndash;1092. https://doi.org/10.1007/s40593-021-00285-9\u003c/li\u003e\n\u003cli\u003eBender, E. M., Gebru, T., McMillan-Major, A., \u0026amp; Shmitchell, S. (2021). On the dangers of stochastic parrots: Can language models be too big? \u003cem\u003eFAccT 2021 - Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency\u003c/em\u003e, 610\u0026ndash;623. https://doi.org/10.1145/3442188.3445922\u003c/li\u003e\n\u003cli\u003eBogina, V., Hartman, A., Kuflik, T., \u0026amp; Shulner-Tal, A. (2022). Educating Software and AI Stakeholders About Algorithmic Fairness, Accountability, Transparency and Ethics. \u003cem\u003eInternational Journal of Artificial Intelligence in Education\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(3), 808\u0026ndash;833. https://doi.org/10.1007/s40593-021-00248-0\u003c/li\u003e\n\u003cli\u003eBozkurt, A. (2023). Generative artificial intelligence (AI) powered conversational educational agents: The inevitable paradigm shift. \u003cem\u003eAsian Journal of Distance Education\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e(1), 2023. https://www.asianjde.com/ojs/index.php/AsianJDE/article/view/718\u003c/li\u003e\n\u003cli\u003eBraun, V., \u0026amp; Clarke, V. (2006). Using thematic analysis in psychology. \u003cem\u003eQualitative Research in Psychology\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e(2), 77\u0026ndash;101. https://doi.org/10.1191/1478088706qp063oa\u003c/li\u003e\n\u003cli\u003eBrynjolfsson, E., \u0026amp; McAfee, A. (2017). \u003cem\u003eArtificial intelligence, for real\u003c/em\u003e. Harvard Business Review Press.\u003c/li\u003e\n\u003cli\u003eBuhmann, A., \u0026amp; Fieseler, C. (2021). Towards a deliberative framework for responsible innovation in artificial intelligence. \u003cem\u003eTechnology in Society\u003c/em\u003e, \u003cem\u003e64\u003c/em\u003e. https://doi.org/10.1016/j.techsoc.2020.101475\u003c/li\u003e\n\u003cli\u003eCamar\u0026eacute;na, S. (2021). Engaging with artificial intelligence (AI) with a bottom-up approach for the purpose of sustainability: Victorian farmers market association, Melbourne Australia. \u003cem\u003eSustainability (Switzerland)\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(16). https://doi.org/10.3390/su13169314\u003c/li\u003e\n\u003cli\u003eCantarini, P. (2022). GOVERNANCE OF AI ALGORITHMS-AN APPROACH FROM EPISTEMOLOGIES OF THE SOUTH. \u003cem\u003eRevista Juridica\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(72), 393\u0026ndash;423. https://doi.org/10.26668/revistajur.2316-753X.v5i72.6286\u003c/li\u003e\n\u003cli\u003eCasillo, M., Cecere, L., Dembele, S. P., Lorusso, A., Santaniello, D., \u0026amp; Valentino, C. (2024). The Metaverse and Revolutionary Perspectives for the Smart Cities of the Future. In X.-S. Yang, S. Sherratt, N. Dey, \u0026amp; A. Joshi (Eds.), \u003cem\u003eLecture Notes in Networks and Systems: Vol. 1004 LNNS\u003c/em\u003e (pp. 215\u0026ndash;225). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-97-3305-7_17\u003c/li\u003e\n\u003cli\u003eCastelli, T., Mocenni, C., \u0026amp; Maria Dimitri, G. (2024). A machine learning approach to assess Sustainable Development Goals food performances: The Italian case. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(1 January). https://doi.org/10.1371/journal.pone.0296465\u003c/li\u003e\n\u003cli\u003eChakraborty, J., Majumder, S., Yu, Z., \u0026amp; Menzies, T. (2020). Fairway: A way to build fair ML software. In P. Devanbu, M. Cohen, \u0026amp; T. Zimmermann (Eds.), \u003cem\u003eESEC/FSE 2020 - Proceedings of the 28th ACM Joint Meeting European Software Engineering Conference and Symposium on the Foundations of Software Engineering\u003c/em\u003e (pp. 654\u0026ndash;665). Association for Computing Machinery, Inc. https://doi.org/10.1145/3368089.3409697\u003c/li\u003e\n\u003cli\u003eCorredor Garc\u0026iacute;a, M. P., Quintero Rengifo, E. F., Ariza Ruiz, E. D., \u0026amp; Pi\u0026ntilde;eros, M. I. A. (2021). Methodological Tools Design to Teach Ethics in Accounting According to IFAC IES 4. An Approximation to the Giving Voice to Values (GVV) Methodology. In M. Botto-Tobar, H. Cruz, \u0026amp; A. D\u0026iacute;az Cadena (Eds.), \u003cem\u003eAdvances in Intelligent Systems and Computing: Vol. 1327 AISC\u003c/em\u003e (pp. 17\u0026ndash;28). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-68083-1_2\u003c/li\u003e\n\u003cli\u003eCrawford, K. (2021). Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. In \u003cem\u003eAtlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence\u003c/em\u003e. https://doi.org/10.56315/pscf3-22crawford\u003c/li\u003e\n\u003cli\u003eDennehy, D., Griva, A., Pouloudi, N., Dwivedi, Y. K., Dwivedi, Y. K., Pappas, I., Pappas, I., \u0026amp; Mantymaki, M. (Eds.). (2021). 20th IFIP WG 6.11 Conference on e-Business, e-Services and e-Society, I3E 2021. \u003cem\u003eLecture Notes in Computer Science\u003c/em\u003e, \u003cem\u003e12896 LNCS\u003c/em\u003e. https://www.scopus.com/inward/record.uri?eid=2-s2.0-85115191822\u0026amp;partnerID=40\u0026amp;md5=dd33c6f1dc80f55f3f9baa7b5865bb76\u003c/li\u003e\n\u003cli\u003eDenyer, D., Tranfield, D., \u0026amp; van Aken, J. E. (2008). Developing Design Propositions through Research Synthesis. \u003cem\u003eOrganization Studies\u003c/em\u003e, \u003cem\u003e29\u003c/em\u003e(3), 393\u0026ndash;413. https://doi.org/10.1177/0170840607088020\u003c/li\u003e\n\u003cli\u003eErman, E., \u0026amp; Furendal, M. (2022). The Global Governance of Artificial Intelligence: Some Normative Concerns. \u003cem\u003eMoral Philosophy and Politics\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(2), 267\u0026ndash;291. https://doi.org/10.1515/mopp-2020-0046\u003c/li\u003e\n\u003cli\u003eEss, C., Kreps, D., Kimppa, K., \u0026amp; Leenen, L. (Eds.). (2018). 13th IFIP TC 9 International Conference on Human Choice and Computers, HCC 2018 Held at the 24th IFIP World Computer Congress, WCC 2018. \u003cem\u003eIFIP Advances in Information and Communication Technology\u003c/em\u003e, \u003cem\u003e537\u003c/em\u003e. https://www.scopus.com/inward/record.uri?eid=2-s2.0-85053888611\u0026amp;partnerID=40\u0026amp;md5=bb8edacf59336970cb7a84e8f049db24\u003c/li\u003e\n\u003cli\u003eEynon, R. (2022). DATAFICATION AND THE ROLE OF SCHOOLING: Challenging the status quo. \u003cem\u003eLearning to Live with Datafication: Educational Case Studies and Initiatives from Across the World\u003c/em\u003e, 17\u0026ndash;34. https://doi.org/10.4324/9781003136842-2\u003c/li\u003e\n\u003cli\u003eFjeld, J., Nagy, A., Hilligoss, H., Achten, N., \u0026amp; Srikumar, M. (2020). Principled Artificial Intelligence: Mapping Consensus in Ethical and Rights-Based Approaches to Principles for AI. \u003cem\u003eSSRN Electronic Journal\u003c/em\u003e. https://doi.org/10.2139/ssrn.3518482\u003c/li\u003e\n\u003cli\u003eFloridi, L. (2014). \u003cem\u003eThe Fourth Revolution: How the Infosphere is Reshaping Human Reality\u003c/em\u003e. OUP Oxford. https://books.google.co.id/books?id=hOF_AwAAQBAJ\u003c/li\u003e\n\u003cli\u003eFloridi, L., \u0026amp; Cowls, J. (2021). A Unified Framework of Five Principles for AI in Society. In \u003cem\u003ePhilosophical Studies Series\u003c/em\u003e (Vol. 144, pp. 5\u0026ndash;17). https://doi.org/10.1007/978-3-030-81907-1_2\u003c/li\u003e\n\u003cli\u003eFloridi, L., Cowls, J., Beltrametti, M., Chatila, R., Chazerand, P., Dignum, V., Luetge, C., Madelin, R., Pagallo, U., Rossi, F., Schafer, B., Valcke, P., \u0026amp; Vayena, E. (2018). AI4People\u0026mdash;An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations. \u003cem\u003eMinds and Machines\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(4), 689\u0026ndash;707. https://doi.org/10.1007/s11023-018-9482-5\u003c/li\u003e\n\u003cli\u003eFrischknecht-Gruber, C. M.-L., Reif, M., \u0026amp; Senn, C. W. (2022). \u003cem\u003eGlobal Efforts Towards Establishing Safety Directives for Intelligent Systems: Review\u003c/em\u003e (M. C. Leva, E. Patelli, L. Podofillini, \u0026amp; S. Wilson (Eds.); pp. 2420\u0026ndash;2427). Research Publishing. https://doi.org/10.3850/978-981-18-5183-4_S11-11-472-cd\u003c/li\u003e\n\u003cli\u003eGutierrez, S. S. M., Perez, S. L., \u0026amp; Munguia, M. G. (2022). Artificial Intelligence in e-Learning Plausible Scenarios in Latin America and New Graduation Competencies. \u003cem\u003eRevista Iberoamericana de Tecnologias Del Aprendizaje\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(1), 31\u0026ndash;40. https://doi.org/10.1109/RITA.2022.3149833\u003c/li\u003e\n\u003cli\u003eGutierrez Y Restrepo, E., \u0026amp; Floris, C. (2022). \u003cem\u003eSupporting Teachers and Students Through a Smart Integrated System for Truly Inclusive Higher Education\u003c/em\u003e. 107\u0026ndash;113. https://doi.org/10.1145/3563137.3563173\u003c/li\u003e\n\u003cli\u003eHelberger, N., Araujo, T., \u0026amp; de Vreese, C. H. (2020). Who is the fairest of them all? Public attitudes and expectations regarding automated decision-making. \u003cem\u003eComputer Law and Security Review\u003c/em\u003e, \u003cem\u003e39\u003c/em\u003e. https://doi.org/10.1016/j.clsr.2020.105456\u003c/li\u003e\n\u003cli\u003eHolmes, W., Porayska-Pomsta, K., Holstein, K., Sutherland, E., Baker, T., Shum, S. B., Santos, O. C., Rodrigo, M. T., Cukurova, M., Bittencourt, I. I., \u0026amp; Koedinger, K. R. (2022). Ethics of AI in Education: Towards a Community-Wide Framework. \u003cem\u003eInternational Journal of Artificial Intelligence in Education\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(3), 504\u0026ndash;526. https://doi.org/10.1007/s40593-021-00239-1\u003c/li\u003e\n\u003cli\u003eJeremy Pitt, J., DIaconescu, A., \u0026amp; Ober, J. (2019). Knowledge Management for Democratic Governance of Socio-Technical Systems. In \u003cem\u003eLecture Notes in Computer Science: Vol. 11300 LNCS\u003c/em\u003e (pp. 38\u0026ndash;61). Springer Verlag [email protected]. https://doi.org/10.1007/978-3-030-05333-8_4\u003c/li\u003e\n\u003cli\u003eJobin, A., Ienca, M., \u0026amp; Vayena, E. (2019). The global landscape of AI ethics guidelines. \u003cem\u003eNature Machine Intelligence\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e(9), 389\u0026ndash;399. https://doi.org/10.1038/s42256-019-0088-2\u003c/li\u003e\n\u003cli\u003eK\u0026auml;stner, C., \u0026amp; Kang, E. (2020). Teaching software engineering for AI-enabled systems. \u003cem\u003eProceedings - International Conference on Software Engineering\u003c/em\u003e, 45\u0026ndash;48. https://doi.org/10.1145/3377814.3381714\u003c/li\u003e\n\u003cli\u003eKeeble, J., \u0026amp; Blatchly-Lewis, M. (2022). \u003cem\u003eThe responsible city\u003c/em\u003e (pp. 139\u0026ndash;154). Wiley Blackwell. https://doi.org/10.1002/9781119746294.ch8\u003c/li\u003e\n\u003cli\u003eKewalramani, S., Kidman, G., \u0026amp; Palaiologou, I. (2021). Using Artificial Intelligence (AI)-interfaced robotic toys in early childhood settings: a case for children\u0026rsquo;s inquiry literacy. \u003cem\u003eEuropean Early Childhood Education Research Journal\u003c/em\u003e, \u003cem\u003e29\u003c/em\u003e(5), 652\u0026ndash;668. https://doi.org/10.1080/1350293X.2021.1968458\u003c/li\u003e\n\u003cli\u003eKhoury, P., Srinivasan, R., Kakumanu, S., Ochoa, S., Keswani, A., Sparks, R., \u0026amp; Rider, N. L. (2022). A Framework for Augmented Intelligence in Allergy and Immunology Practice and Research\u0026mdash;A Work Group Report of the AAAAI Health Informatics, Technology, and Education Committee. \u003cem\u003eJournal of Allergy and Clinical Immunology: In Practice\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(5), 1178\u0026ndash;1188. https://doi.org/10.1016/j.jaip.2022.01.047\u003c/li\u003e\n\u003cli\u003eKlerkx, L., \u0026amp; Rose, D. (2020). Dealing with the game-changing technologies of Agriculture 4.0: How do we manage diversity and responsibility in food system transition pathways? \u003cem\u003eGlobal Food Security\u003c/em\u003e, \u003cem\u003e24\u003c/em\u003e. https://doi.org/10.1016/j.gfs.2019.100347\u003c/li\u003e\n\u003cli\u003eKnox, J. (2020). Artificial intelligence and education in China. \u003cem\u003eLearning, Media and Technology\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e(3), 298\u0026ndash;311. https://doi.org/10.1080/17439884.2020.1754236\u003c/li\u003e\n\u003cli\u003eKritt, D. W., \u0026amp; Winegar, L. T. (2007). \u003cem\u003eEducation and technology : critical perspectives, possible futures\u003c/em\u003e. Lexington Books.\u003c/li\u003e\n\u003cli\u003eLeslie, D. (2019). \u003cem\u003eUnderstanding artificial intelligence ethics and safety\u003c/em\u003e. https://doi.org/10.5281/zenodo.3240529\u003c/li\u003e\n\u003cli\u003eLosavio, M. (2021). Algorithms of Machines and Law: Risks in Pattern Recognition, Machine Learning and Artificial Intelligence for Justice and Fairness. \u003cem\u003ePublic Governance, Administration and Finances Law Review\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(2), 21\u0026ndash;34. https://doi.org/10.53116/pgaflr.2021.2.3\u003c/li\u003e\n\u003cli\u003eLuckin, R. (2018). Machine Learning and Human Intelligence: The Future of Education for the 21st Century. In \u003cem\u003eUCL IOE Press\u003c/em\u003e. UCL IOE Press. https://eric.ed.gov/?id=ED584840\u003c/li\u003e\n\u003cli\u003eLuetge, C., Poszler, F., Acosta, A. J., Danks, D., Gottehrer, G., Mihet-Popa, L., \u0026amp; Naseer, A. (2021). AI4people: Ethical guidelines for the automotive sector-fundamental requirements and practical recommendations. \u003cem\u003eInternational Journal of Technoethics\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(1), 101\u0026ndash;125. https://doi.org/10.4018/IJT.20210101.oa2\u003c/li\u003e\n\u003cli\u003eMcKay, F., Williams, B. J., Prestwich, G., Bansal, D., Hallowell, N., \u0026amp; Treanor, D. (2022). The ethical challenges of artificial intelligence-driven digital pathology. \u003cem\u003eJournal of Pathology: Clinical Research\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(3), 209\u0026ndash;216. https://doi.org/10.1002/cjp2.263\u003c/li\u003e\n\u003cli\u003eMethley, A. M., Campbell, S., Chew-Graham, C., McNally, R., \u0026amp; Cheraghi-Sohi, S. (2014). PICO, PICOS and SPIDER: A comparison study of specificity and sensitivity in three search tools for qualitative systematic reviews. \u003cem\u003eBMC Health Services Research\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e(1). https://doi.org/10.1186/s12913-014-0579-0\u003c/li\u003e\n\u003cli\u003eO\u0026rsquo;Connor, C., \u0026amp; Joffe, H. (2020). Intercoder Reliability in Qualitative Research: Debates and Practical Guidelines. \u003cem\u003eInternational Journal of Qualitative Methods\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e. https://doi.org/10.1177/1609406919899220\u003c/li\u003e\n\u003cli\u003eOECD. (2022). \u003cem\u003eOECD, Recommendation of the Council on Artificial Intelligence, OECD/LEGAL/0449\u003c/em\u003e. http://legalinstruments.oecd.org\u003c/li\u003e\n\u003cli\u003ePage, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hr\u0026oacute;bjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., \u0026hellip; Moher, D. (2021). The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. \u003cem\u003eBMJ\u003c/em\u003e, n71. https://doi.org/10.1136/bmj.n71\u003c/li\u003e\n\u003cli\u003ePaic, A. (2019). Policies for artificial intelligence in science and innovation. \u003cem\u003eProceedings of Science\u003c/em\u003e, \u003cem\u003e372\u003c/em\u003e. https://www.scopus.com/inward/record.uri?eid=2-s2.0-85101244456\u0026amp;partnerID=40\u0026amp;md5=9a320da7b8f618b5b7a690bd0f6e2fee\u003c/li\u003e\n\u003cli\u003ePopay, J., Roberts, H., Sowden, A., Petticrew, M., Arai, L., Rodgers, M., \u0026amp; et all. (2006). Guidance on the conduct of narrative synthesis in systematic reviews. a product from the a product from the ESRC methods programme. \u003cem\u003eEsrc\u003c/em\u003e, \u003cem\u003eversion 1\u003c/em\u003e(April 2006), 1\u0026ndash;92. https://doi.org/10.13140/2.1.1018.4643\u003c/li\u003e\n\u003cli\u003ePorter, M. E., \u0026amp; Kramer, M. R. (2011). Harvard Business Riview. \u003cem\u003eHarvard Business Review\u003c/em\u003e, \u003cem\u003e89\u003c/em\u003e(1\u0026ndash;2).\u003c/li\u003e\n\u003cli\u003ePrice, W. N., \u0026amp; Cohen, I. G. (2019). Privacy in the age of medical big data. \u003cem\u003eNature Medicine\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(1), 37\u0026ndash;43. https://doi.org/10.1038/s41591-018-0272-7\u003c/li\u003e\n\u003cli\u003eRajagopal, N. K., Anand, M., \u0026amp; Mohanty, S. (2024). Exploring Machine Learning Applications in Human Resources Management: A Comprehensive Review. In \u003cem\u003eStudies in Systems, Decision and Control\u003c/em\u003e (Vol. 569, pp. 303\u0026ndash;313). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-71649-2_26\u003c/li\u003e\n\u003cli\u003eRocco, R., Thomas, A., \u0026amp; Novas-Ferrad\u0026aacute;s, M. (2022). \u003cem\u003eTeaching design for values: Concepts, tools \u0026amp; practices\u003c/em\u003e. TU Delft. https://doi.org/10.34641/mg.54\u003c/li\u003e\n\u003cli\u003eRousseau, D. M. (2006). Is there Such a thing as \u0026ldquo;Evidence-Based Management\u0026rdquo;? \u003cem\u003eAcademy of Management Review\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e(2), 256\u0026ndash;269. https://doi.org/10.5465/amr.2006.20208679\u003c/li\u003e\n\u003cli\u003eSapienza, S. (2022). Current Trends, Machine Learning, and Food Safety Data Governance. In \u003cem\u003eLaw, Governance and Technology Series\u003c/em\u003e (Vol. 52, pp. 123\u0026ndash;160). Springer Nature. https://doi.org/10.1007/978-3-031-09367-8_4\u003c/li\u003e\n\u003cli\u003eSchiff, D., Rakova, B., Ayesh, A., Fanti, A., \u0026amp; Lennon, M. (2021). Explaining the Principles to Practices Gap in AI. \u003cem\u003eIEEE Technology and Society Magazine\u003c/em\u003e, \u003cem\u003e40\u003c/em\u003e(2), 81\u0026ndash;94. https://doi.org/10.1109/MTS.2021.3056286\u003c/li\u003e\n\u003cli\u003eSchiff, D. S., Laas, K., Biddle, J. B., \u0026amp; Borenstein, J. (2022). Global AI Ethics Documents: What They Reveal About Motivations, Practices, and Policies. In \u003cem\u003eInternational Library of Ethics, Law and Technology\u003c/em\u003e (Vol. 23, pp. 121\u0026ndash;143). Springer Science and Business Media B.V. https://doi.org/10.1007/978-3-030-86201-5_7\u003c/li\u003e\n\u003cli\u003eSelwyn, N. (2019). Should Robots Replace Teachers? AI and the Future of Education. In \u003cem\u003eJohn Wiley \u0026amp; Sons, Inc\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eSelwyn, N. (2021a). Ed-Tech Within Limits: Anticipating educational technology in times of environmental crisis. \u003cem\u003eE-Learning and Digital Media\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e(5), 496\u0026ndash;510. https://doi.org/10.1177/20427530211022951\u003c/li\u003e\n\u003cli\u003eSelwyn, N. (2021b). \u003cem\u003eEducation and technology: Key issues and debates\u003c/em\u003e. Bloomsbury Publishing.\u003c/li\u003e\n\u003cli\u003eSen, A. (2017). Development as Freedom. In \u003cem\u003eThe Top 50 Sustainability Books\u003c/em\u003e. Oxford University Press. https://doi.org/10.4324/9781351279086-33\u003c/li\u003e\n\u003cli\u003eShah, C. (2023). Generative AI and the Future of Information Access. \u003cem\u003eUNESCO\u003c/em\u003e, \u003cem\u003eJuly\u003c/em\u003e, 3\u0026ndash;3. https://doi.org/10.1145/3583780.3615317\u003c/li\u003e\n\u003cli\u003eSharma, S., Rawal, Y. S., Pal, S., \u0026amp; Dani, R. (2022). Fairness, Accountability, Sustainability, Transparency (FAST) of Artificial Intelligence in Terms of Hospitality Industry. In S. Fong, N. Dey, \u0026amp; A. Joshi (Eds.), \u003cem\u003eLecture Notes in Networks and Systems\u003c/em\u003e (Vol. 314, pp. 495\u0026ndash;504). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-16-5655-2_48\u003c/li\u003e\n\u003cli\u003eShneiderman, B. (2020). Bridging the gap between ethics and practice: Guidelines for reliable, safe, and trustworthy human-centered AI systems. \u003cem\u003eACM Transactions on Interactive Intelligent Systems\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(4). https://doi.org/10.1145/3419764\u003c/li\u003e\n\u003cli\u003eSijing, L., \u0026amp; Lan, W. (2018). Artificial intelligence education ethical problems and solutions. \u003cem\u003e13th International Conference on Computer Science and Education, ICCSE 2018\u003c/em\u003e, 155\u0026ndash;158. https://doi.org/10.1109/ICCSE.2018.8468773\u003c/li\u003e\n\u003cli\u003eSpiegel, J. M., Wilkinson, J., Ehrlich, R., Lockhart, K., Yassi, A., Barker, S., Riera, F., \u0026amp; Kistnasamy, B. (2021). Using artificial intelligence for high-volume identification of silicosis and tuberculosis a bio-ethics approach. \u003cem\u003eAnnals of Global Health\u003c/em\u003e, \u003cem\u003e87\u003c/em\u003e(1). https://doi.org/10.5334/aogh.3206\u003c/li\u003e\n\u003cli\u003eStehr, N. (2017). Knowledge Societies. In \u003cem\u003eSociety \u0026amp; Knowledge\u003c/em\u003e (pp. 299\u0026ndash;322). Routledge. https://doi.org/10.4324/9781315129884-13\u003c/li\u003e\n\u003cli\u003eSu, A. (2022). The Promise and Perils of International Human Rights Law for AI Governance. \u003cem\u003eLaw, Technology and Humans\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(1). https://doi.org/10.5204/LTHJ.2332\u003c/li\u003e\n\u003cli\u003eTanskanen, K., Ahola, T., Aminoff, A., Bragge, J., Kaipia, R., \u0026amp; Kauppi, K. (2017). Towards evidence-based management of external resources: Developing design propositions and future research avenues through research synthesis. \u003cem\u003eResearch Policy\u003c/em\u003e, \u003cem\u003e46\u003c/em\u003e(6), 1087\u0026ndash;1105. https://doi.org/10.1016/j.respol.2017.04.002\u003c/li\u003e\n\u003cli\u003eTranfield, D., Denyer, D., \u0026amp; Smart, P. (2003). Towards a Methodology for Developing Evidence‐Informed Management Knowledge by Means of Systematic Review. \u003cem\u003eBritish Journal of Management\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e(3), 207\u0026ndash;222. https://doi.org/10.1111/1467-8551.00375\u003c/li\u003e\n\u003cli\u003eUNESCO. (2005). \u003cem\u003e2005 Convention on Diversity of Cultural Expressions\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eUNESCO. (2019). Global Education Monitoring Report 2019: Migration, Displacement and Education \u0026ndash; Building Bridges, not Walls. \u003cem\u003eGlobal Education Monitoring Report 2019: Migration, Displacement and Education \u0026ndash; Building Bridges, Not Walls\u003c/em\u003e. https://doi.org/10.54676/xdzd4287\u003c/li\u003e\n\u003cli\u003eUNESCO. (2021a). AI and education: guidance for policy-makers. In \u003cem\u003eAI and education: guidance for policy-makers\u003c/em\u003e. UNESCO Publishing. https://doi.org/10.54675/pcsp7350\u003c/li\u003e\n\u003cli\u003eUNESCO. (2021b). \u003cem\u003ePublications\u003c/em\u003e (Vol. 17). http://unesdoc.unesco.org/\u003c/li\u003e\n\u003cli\u003eUNESCO. (2023). Global Education Monitoring Report 2023: Technology in education: A tool on whose terms? In \u003cem\u003eGlobal Education Monitoring Report 2023: Technology in education: A tool on whose terms?\u003c/em\u003e GEM Report UNESCO. https://doi.org/10.54676/uzqv8501\u003c/li\u003e\n\u003cli\u003eVon Schomberg, R. (2013). A vision of responsible research and innovation. Responsible Innovation: Managing the Responsible Emergence of Science and Innovation in Society (Owen, R., Bessant, J. and Heintz, M., eds). \u003cem\u003eResponsible Innovation\u003c/em\u003e, 51\u0026ndash;74.\u003c/li\u003e\n\u003cli\u003eWeber-Lewerenz, B. (2021). Corporate digital responsibility (CDR) in construction engineering\u0026mdash;ethical guidelines for the application of digital transformation and artificial intelligence (AI) in user practice. \u003cem\u003eSN Applied Sciences\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e(10). https://doi.org/10.1007/s42452-021-04776-1\u003c/li\u003e\n\u003cli\u003eWhittlestone, J., Nyrup, R., Alexandrova, A., Dihal, K., \u0026amp; Cave, S. (2019). Ethical and societal implications of algorithms, data, and artificial intelligence: a roadmap for research. \u003cem\u003eLondon: Nuffield \u0026hellip;\u003c/em\u003e, 1\u0026ndash;59. https://www.nuffieldfoundation.org/wp-content/uploads/2019/02/Ethical-and-Societal-Implications-of-Data-and-AI-report-Nuffield-Foundat.pdf%0Ahttp://www.nuffieldfoundation.org/sites/default/files/files/Ethical-and-Societal-Implications-of-Data-and-AI-repor\u003c/li\u003e\n\u003cli\u003eWilliamson, B. (2024). The Social life of AI in Education. \u003cem\u003eInternational Journal of Artificial Intelligence in Education\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e(1), 97\u0026ndash;104. https://doi.org/10.1007/s40593-023-00342-5\u003c/li\u003e\n\u003cli\u003eWilliamson, B., Bayne, S., \u0026amp; Shay, S. (2020). The datafication of teaching in Higher Education: critical issues and perspectives. \u003cem\u003eTeaching in Higher Education\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(4), 351\u0026ndash;365. https://doi.org/10.1080/13562517.2020.1748811\u003c/li\u003e\n\u003cli\u003eWilliamson, B., \u0026amp; Eynon, R. (2020). Historical threads, missing links, and future directions in AI in education. \u003cem\u003eLearning, Media and Technology\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e(3), 223\u0026ndash;235. https://doi.org/10.1080/17439884.2020.1798995\u003c/li\u003e\n\u003cli\u003eXu, L. (2020). The Dilemma and Countermeasures of AI in Educational Application. \u003cem\u003eACM International Conference Proceeding Series\u003c/em\u003e, 289\u0026ndash;294. https://doi.org/10.1145/3445815.3445863\u003c/li\u003e\n\u003cli\u003eZawacki-Richter, O., Mar\u0026iacute;n, V. I., Bond, M., \u0026amp; Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education \u0026ndash; where are the educators? \u003cem\u003eInternational Journal of Educational Technology in Higher Education\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(1), 39. https://doi.org/10.1186/s41239-019-0171-0\u003c/li\u003e\n\u003cli\u003eZhang, K., \u0026amp; Aslan, A. B. (2021). AI technologies for education: Recent research \u0026amp;amp; future directions. \u003cem\u003eComputers and Education: Artificial Intelligence\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e, 100025. https://doi.org/10.1016/j.caeai.2021.100025\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Ethical AI, Inclusive education, Pedagogically-Centered AI Governance (PCAG), Systematic review, Educational ethics, AI governance.","lastPublishedDoi":"10.21203/rs.3.rs-8230657/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8230657/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003ePurpose:\u003c/h2\u003e \u003cp\u003eTo bridge the persistent gap between AI ethics principles and educational practice by proposing a \u003cem\u003ePedagogically-Centered AI Governance\u003c/em\u003e (PCAG) model that situates pedagogy as the generative core of ethical and inclusive artificial intelligence.\u003c/p\u003e\u003ch2\u003eDesign/methodology/approach:\u003c/h2\u003e \u003cp\u003eA systematic literature review was conducted following PRISMA 2020 guidelines across seven databases (Scopus, Web of Science, IEEE Xplore, ScienceDirect, SpringerLink, Wiley, and Google Scholar) covering publications from 2014 to May 2024. Using the \u003cb\u003ePICo\u003c/b\u003e framework for scoping and \u003cb\u003eCIMO\u003c/b\u003e logic for synthesis, \u003cb\u003e127\u003c/b\u003e peer-reviewed studies were analyzed to identify mechanisms linking ethical AI to inclusive educational outcomes.\u003c/p\u003e\u003ch2\u003eFindings:\u003c/h2\u003e \u003cp\u003eFour recurrent mechanisms were synthesized from the literature: (1) \u003cem\u003epedagogical capacity-building\u003c/em\u003e, (2) \u003cem\u003etechnical process control\u003c/em\u003e, (3) \u003cem\u003eregulatory accountability\u003c/em\u003e, and (4) \u003cem\u003eparticipatory legitimation\u003c/em\u003e. Integrating these mechanisms, the study proposes the \u003cb\u003ePedagogically-Centered AI Governance (PCAG)\u003c/b\u003e model and the \u003cb\u003eSustainable Educational Governance Framework (SEG-F)\u003c/b\u003e. Together, they conceptualize AI not merely as a technological tool but as an \u003cem\u003eeducational infrastructure\u003c/em\u003e that should be ethically and pedagogically governed to promote equity, inclusion, and epistemic justice.\u003c/p\u003e\u003ch2\u003ePractical implications:\u003c/h2\u003e \u003cp\u003ePCAG offers a multi-layered governance architecture and a six-item operational checklist for policymakers, educators, and designers to translate AI ethics into curriculum development, teacher training, and institutional decision-making.\u003c/p\u003e\u003ch2\u003eOriginality/value:\u003c/h2\u003e \u003cp\u003eThis review advances AI-in-education scholarship by combining \u003cb\u003ePRISMA\u0026thinsp;+\u0026thinsp;PICo\u0026thinsp;+\u0026thinsp;CIMO\u003c/b\u003e into a mechanism-oriented synthesis and by centering pedagogy as the key lever for converting global AI ethics principles into contextually just and inclusive educational practices.\u003c/p\u003e","manuscriptTitle":"Ethical and Inclusive Artificial Intelligence for Sustainable Knowledge Societies: A Global Governance Framework through a Systematic Literature Review","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-24 08:23:06","doi":"10.21203/rs.3.rs-8230657/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"0bc364f6-d933-410b-a049-38261ab2f95b","owner":[],"postedDate":"December 24th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-03-19T08:57:01+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-24 08:23:06","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8230657","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8230657","identity":"rs-8230657","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}