Defining AI Literacy in Higher Education

Artificial Intelligence Literacy (AIL) represents a set of competencies that empower individuals to understand, engage with, and critically evaluate AI technologies. In higher education, this necessitates a comprehensive framework that integrates technical proficiency, ethical reasoning, collaborative skills, emotional intelligence, and sociocultural awareness. By integrating AIL taxonomies presented in recent educational research reviews (Rizvi et al., 2023; Shiri, 2024), the AIL framework proposed by Pretorius and Cahusac de Caux (2024), can be expanded to include seven domains: foundational, conceptual, ethical, social, affective, sociocultural, and pedagogical (figure 1 below).

The foundational domain provides technical knowledge of AI systems, including their capabilities, limitations, and underlying mechanisms. It forms the basis for engaging with AI technologies effectively, mirroring the importance of foundational knowledge highlighted in digital literacy frameworks (Shiri, 2024).

Building on this, the conceptual domain emphasizes critical evaluation and the ability to apply AI concepts effectively within diverse academic and professional contexts. Together, these dimensions ensure that learners acquire both the technical and cognitive foundations required to navigate the complexities of AI. (Ng et al., 2021).

The ethical domain focuses on principles such as bias, fairness, accountability, inclusivity, privacy in AI use (Kharkova & Holmes, 2022). This domain aligns with the call for ethical and responsible AI practices in emerging AI literacy research (Long & Magerko, 2020; Rizvi et al., 2023).

Complementing this is the social domain, which focuses on collaboration and human-AI interactions. This dimension equips learners to work effectively individually or within interdisciplinary teams and to utilize AI as a tool for enhancing collective problem-solving and innovation.

The affective domain introduces an essential element of emotional engagement, recognizing that learners' attitudes toward AI significantly influence their openness to AI technology adoption and critical reflection. By supporting resilience, adaptability, and positive emotional responses, AIL programmes sustain learners to navigate the layered challenges posed by the use of AI technologies (Gillespie et al., 2023).

Expanding on the above five core AIL domains proposed by Pretorius and Cahusac de Caux (2024), two additional domains are proposed:

The sociocultural domain addresses the broader societal, cultural, and global contexts in which AI operates. Drawing from digital literacies and multiliteracies theories (Bawden, 2008; Cazden et al, 1996), this dimension emphasizes the importance of ethical and sustainable AI practices that align with global challenges such as climate change and equity. It also highlights the role of AI in shaping cultural norms and governance structures, advocating for inclusivity in AI design and deployment (Holmes, 2023). By encouraging learners to critically evaluate AI’s societal impact, the sociocultural dimension prepares them for the future of work and to engage with AI in ways that promote social justice.

The pedagogical domain focuses on empowering educators with strategies to effectively facilitate AI literacy teaching and learning activities and adopt AI use in educational contexts (Kong et al., 2023; Shiri, 2024). It emphasizes educator training and interdisciplinary integration, promoting teaching strategies that adapt to diverse learning contexts (Laupichler et al., 2022). Drawing upon approaches to digital literacy education (Meyers et al., 2013), it encompasses both formal and informal learning strategies. Furthermore, it supports professional development for educators, equipping them with the tools and resources needed for AIL curriculum integration.

This expanded AIL framework is intended to also reflect the participatory and adaptive ethos of digital literacies (Meyers et al., 2013), emphasizing critical engagement, contextual application, and cultural inclusivity; in particular, AIL programmes should promote:

• Reflexivity: Encouraging learners to critically assess their relationship with AI technologies and their societal roles.

• Adaptability: Recognizing the need to review conceptual understanding and approaches in response to the rapid evolution of AI technologies and their applications.

By synthesizing the range of AIL taxonomies presented in the literature and incorporating the pedagogical dimension, the expanded AIL framework provides an analytical tool to evaluate current AIL initiatives in higher education for the scope of this comparative study.

Cultural Change in Universities

Achieving high levels of multi-domain AIL across a higher education institution requires a profound cultural shift among students and faculty. Universities are complex organizations with ingrained traditions, values, and practices. The widespread introduction of AI technologies challenges existing pedagogical approaches and necessitates adaptation to new ways of teaching, learning, and researching. Cultural archetypes within institutions influence how change is perceived and enacted, suggesting that a one-size-fits-all approach is ineffective. Hargreaves et al. (1998) highlight that successful innovation depends on acknowledging the multifaceted nature of educational institutions. Factors such as organizational innovations, managed or planned educational change, and mandated reforms play critical roles in how higher education adapts to new technologies.

Kezar and Eckel (2000) emphasize the significant impact of institutional culture on change strategies, underscoring the need to understand unique institutional cultures when implementing comprehensive changes. The integration of AI technologies requires universities to address not only pedagogical adjustments but also the broader implications for the future of work and discipline-specific applications of AI. Faculty and students must engage with the socio-cultural dimension of AI not just as a tool but as a transformative force that influences all aspects of academia and society.

A Multi-Level Approach to AI Literacy

Addressing the challenges of integrating AIL programmes requires a multi-level approach operating both institution-wide and at the departmental level. Institution-wide initiatives have the potential to foster an AI-aware culture that embraces innovation and ethical considerations, involving shared AI guidelines for educators and students, promoting informal learning opportunities, and encouraging cross-disciplinary collaboration. At the departmental level, curriculum integration contextualizes AI within specific disciplines, tailoring AIL content to allow students to gain relevant and practical competencies.

AIL initiatives for students and educators are being extensively implemented by universities at the institution-wide level to address the new educational landscape, particularly following the widespread adoption of generative AI technologies; on the other hand, department-level initiatives for curriculum integration are not as common outside of STEM departments (Southworth et al., 2023).

Formal learning strategies such as microlearning modules, credentialing programmes, and project-based learning provide foundational knowledge, enhance engagement and deepen conceptual, social and ethical understanding of AI. Informal opportunities, such as engagement with hands-on workshops, debates, public scholarship and media initiatives, address the social and affective dimensions by fostering a community of practice and encouraging reflective dialogue (Lee et al., 2024; Ng, 2021).

Limitations

The study is exploratory and based on secondary research from publicly available web-based sources, which may not fully represent the scope of each programme due to limited depth of analysis for each individual case.

Moreover, variations in institutional culture and regional educational policies may affect the generalizability of findings.

Future research could include direct engagement with programme administrators, faculty and students, through qualitative methods to evaluate in depth the content, approach and context of each programme.

Ethical Considerations

The study adhered to ethical research practices by:

• Using public data: Only publicly accessible information was used, respecting institutional privacy.

• Accurate representation: Efforts were made to accurately represent each institution's AIL programme without bias.

• Acknowledging Sources: proper citations were provided for all referenced materials and shared in the annexes.

Emphasis on AI Literacy Dimensions

Most of the AIL programmes analyzed prioritize the foundational and conceptual dimensions, ensuring students gain essential technical knowledge about AI technologies and their applications. Ethical considerations are acknowledged to varying degrees, with the majority highlighting responsible AI use and providing guidelines to promote ethical practices. AI teaching guidance documents and resources published by universities tend to address best practices related to the social dimension of student-AI interaction and collaboration, as well as provide practical cases of classroom use of AI for teaching and learning activities.

However, a significant gap can be inferred in the limited references to the sociocultural and emotional dimensions in the AIL programme content and teaching guidance. Few programmes explicitly engage with learners' emotional responses to AI or promote activities that foster emotional intelligence and critical reflection on AI's societal impact.

Educational Strategies Adopted

The AIL programmes analyzed employ diverse strategies:

• Practical guidance and resource provision: There is widespread use of guides and strategies to assist in integrating AI tools into teaching, learning, and assessment practices.

• Online learning and microlearning: Self-paced courses and microlearning modules provide flexibility and cater to diverse learning needs were among the most common modes of instruction.

• Project-based learning: This approach was limited to few AIL programmes, which engage students in hands-on projects that allow them to apply their knowledge in practical contexts.

• Curriculum integration: Some institutions make AI education mandatory for all students, emphasizing interdisciplinary application. Others incentivize completion of online or on-campus courses through micro-credentialing programmes and certificate programmes.

• Educator professional development: Asynchronous training and resources are also commonly adopted to support educators in effectively teaching AI literacy and integrate it into existing courses.

Emerging Themes, Trends and Gaps

Key themes include a strong focus on technical skills, acknowledgment of ethical implications, limited attention to social and affective dimensions, diverse educational strategies, and challenges in making AI literacy relevant to non-technical fields. While there is an effort to embed AI literacy across disciplines, bridging theoretical concepts with practical applications remains limited.

Trends include institution-wide AI literacy initiatives and resource provision for educators. Offering micro-credentials recognizes and incentivizes student engagement in AI literacy. However, gaps remain in addressing the affective dimension and ensuring engagement in non-technical disciplines.

Recommendations for Future Research and Implementations

Future research should explore in-depth programme evaluations, longitudinal studies on AI literacy impact, development of assessment tools for social and affective competencies, and investigation of institutional culture's influence on AIL initiatives.