Embedding Digital Responsibility: An Exploratory Analysis of AI Embedding Digital Responsibility: An Exploratory Analysis of AI Policies in Higher Education Policies in Higher Education

From innovation to inclusion: Advancing equity through AI policy and governance in South African higher education

As artificial intelligence (AI) becomes increasingly embedded in higher education, empirical evidence on how institutional governance shapes its equitable and responsible implementation in South African universities remains limited. This study examined how institutional policies and governance practices influence the implementation and equitable use of AI in undergraduate education while contributing to international discourse on responsible AI governance. Guided by an interpretivist research paradigm, the study adopted a qualitative approach and employed a single-case study design within one public university in South Africa. Participants comprised university leaders, academic staff, professional support staff, and undergraduate students involved in or affected by AI governance and implementation. Data were collected through semi-structured interviews, focus group discussions, and document analysis and analysed using thematic analysis supported by inductive coding. The findings indicate that limited policy transparency, context-insensitive governance frameworks, unequal access to AI technologies, and weak institutional accountability can reinforce educational inequalities, particularly among first-generation and under-resourced students. Conversely, participatory governance, transparent decision-making, stakeholder engagement, and enhanced digital literacy promote more equitable and responsible AI implementation. The study proposes a multi-level governance model integrating institutional policy, stakeholder participation, and pedagogical practice to strengthen equitable AI adoption. It concludes that higher education institutions should develop context-sensitive AI governance frameworks, strengthen institutional capacity, and expand equitable access to AI technologies to advance fairness, inclusion, and responsible AI implementation.

Bridging Ethics and Regulation: A Conceptual Framework for Governing Generative AI in Higher Education

The rapid adoption of generative artificial intelligence (GenAI) in higher education has outpaced institutional readiness, creating urgent ethical and regulatory challenges that threaten academic integrity, data privacy, and educational equity. While global frameworks like UNESCO’s Guidance for Generative AI in Education (2023) advocate for human-centric design, and national laws such as FERPA mandate student data protection, no existing model systematically integrates these domains into a cohesive governance structure. This study addresses this critical gap by proposing the Ethico-Regulatory Governance (ERG) Framework, a conceptual model designed to bridge global ethics with local compliance. Developed through a systematic synthesis of 68 peer-reviewed studies, policy documents, and institutional guidelines, the ERG Framework consists of four interlocking layers: Foundational Principles (UNESCO values), Regulatory Anchors (FERPA/GDPR alignment), Institutional Mechanisms (audits, disclosure, training), and Pedagogical Integration (process-based assessment, prompt engineering). The framework transforms abstract principles into actionable practices, enabling institutions to move beyond reactive policies toward proactive, accountable governance. Key findings demonstrate that effective AI integration requires not only technical oversight but also stakeholder co-design, bias mitigation, and continuous feedback loops. By operationalizing ethics through enforceable mechanisms, the ERG Framework offers a scalable, adaptable solution for universities navigating the complexities of GenAI. Its implementation can safeguard core academic values while fostering innovation, ensuring that AI serves as a partner—not a replacement—for human judgment in teaching, learning, and research.

Policy Silence as Governance: AI Decision-Making, Responsibility, and Risk in K–12 Education

This study examines how policy silence functions as a governance mechanism in the context of generative artificial intelligence (AI) in K–12 education. While existing research has focused on how districts regulate or integrate AI, far less attention has been given to what happens when formal guidance is delayed, incomplete, or unresolved. Drawing on a single-district qualitative case study of a Texas public school district, the analysis uses interviews and document analysis as primary data sources, while student survey data provide descriptive context regarding patterns of AI access and guidance. Findings show that policy silence actively redistributes interpretive authority to educators and school leaders, shifting responsibility for ethical and instructional decision-making onto individual classrooms without corresponding institutional support. This redistribution produces uneven enactment and may contribute to disparities in student access, guidance, and learning opportunities. Applying Critical Policy Analysis and Jencks’ framework of educational opportunity, the study shows that policy silence is not the absence of governance but a governance choice, one that shapes how access, responsibility, and fairness are determined. Equitable AI integration requires policies that pair clarity with support for professional judgment, positioning AI governance as central to contemporary school improvement.

Generative AI Governance in Higher Education (Systematic Literature Review)

This study aimed to provide a comprehensive review of generative AI governance in higher education during the period 2022–2026, in light of the rapid expansion of the use of big language models and content generation applications within universities. The study adopted a comprehensive scope review methodology, guided by the SALSA framework for the research, evaluation, synthesis, and analysis phases, and by PRISMA-ScR guidelines for documenting the review procedures. The study analyzed peer-reviewed scientific literature, sectoral, institutional, and regulatory frameworks, and selected university policies, with particular attention to the Arab context and research and regulatory gaps. The results showed that generative AI governance is not limited to addressing issues of cheating and plagiarism, but encompasses interconnected dimensions, most notably: academic integrity, privacy, transparency, disclosure, fairness, capacity building, human oversight, risk management, and assessment redesign. The study also indicated that the global trend is moving toward responsible and conditional use rather than outright prohibition or unregulated adoption. The study concluded that effective governance requires combining research evidence, sectoral frameworks, and institutional policies, while translating general principles into clear procedures at the university and course levels. It also revealed the need for more specialized policies that address privacy, linguistic equity, data protection, and the transparency. The study recommends developing flexible university guidelines, disclosure models, data use controls, ongoing training programs, and periodic review mechanisms that ensure a balance between innovation, quality of learning, and academic integrity.

Layered AI ethics and governance: corporate codes and public instruments in the AI era

Trust as a Mediating Mechanism in AI-Enabled School Leadership: Navigating Benefits, Risks, and Ethical Tensions in Education 4.0

Artificial intelligence (AI) is rapidly reshaping school leadership within Education 4.0, offering enhanced decision-making and organisational efficiency while intensifying ethical concerns regarding transparency, bias, and accountability. Existing research has largely treated these opportunities and risks as separate phenomena, overlooking the relational processes through which AI is enacted in practice. This paper advances a process-based conceptualisation by positioning trust as the central mediating mechanism in AI-enabled school leadership. It argues that AI does not produce outcomes directly; rather, its effects are contingent on how it is accepted, interpreted, and enacted within school contexts. The proposed framework shows that trust shapes whether AI leads to constructive outcomes, including ethical use, professional engagement, and improvement, or to disruptive consequences such as resistance and mistrust. Leadership is conceptualised as a key antecedent of trust, highlighting the centrality of relational governance in the effective and responsible integration of AI in schools.