AI as a bridge: How technology facilitates educational and workforce transitions in a digital era

 Corresponding author: Dinesh Deckker Copyright © 2025 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution License 4.0. AI as a bridge: How technology facilitates educational and workforce transitions in a digital era Dinesh Deckker 1, *, Subhashini Sumanasekara 2 and Kadir Uludag 3 1 Faculty of Arts, Science and Technology, Wrexham University, United Kingdom. 2 Faculty of Computing and Social Sciences, University of Gloucestershire, United Kingdom. 3 Faculty Member at the Capital Medical University, Xuanwu Hospital, Department of Neurology, Beijing, China. World Journal of Advanced Research and Reviews, 2025, 26(03), 2491-2504 Publication history: Received on 17 May 2025; revised on 23 June 2025; accepted on 26 June 2025 Article DOI: https://doi.org/10.30574/wjarr.2025.26.3.2458 Abstract As artificial intelligence (AI) becomes deeply embedded in educational and workforce systems, its role as a transformative enabler of educational-to-employment transitions is increasingly significant. This paper discusses how AI functions not merely as a pedagogical tool or employment disruptor, but as a “transitional infrastructure” connecting formal learning environments with the evolving demands of digitally mediated labour markets. The study employs a narrative review methodology, synthesising interdisciplinary literature from 2020 to 2025, drawing on foundational educational theories and recent empirical research. Five key themes emerged: AI-enhanced personalised learning, intelligent career guidance, institutional transformation through smart education, equity in digital inclusion, and pedagogical-ethical considerations. Findings suggest that AI can support lifelong learning, skill adaptation, and systemic reform when implemented ethically and inclusively. However, the review also highlights challenges such as algorithmic bias, data inequality, and the erosion of learner agency. The review calls for equity-centred, evidence-based frameworks to ensure AI fosters inclusive, just, and future-ready learning-to-work pathways. Keywords: Artificial Intelligence in Education; Workforce Transitions; Personalised Learning; Career Readiness; Reskilling and Upskilling; Smart Universities; Educational Equity; Digital Pedagogy; Ethical AI Implementation; Educational Technology Policy 1. Introduction The rise of artificial intelligence (AI) as a transformative force across sectors has raised crucial questions about its role in bridging two historically distinct domains: education and employment. In a rapidly evolving digital economy, traditional linear pathways from schooling to lifelong careers are becoming increasingly obsolete. The accelerating pace of technological innovation, driven by the Fourth Industrial Revolution (4IR), demands flexible, interdisciplinary, and technologically literate individuals (Chaka, 2023). AI stands at the core of this transformation, reshaping not only how individuals learn but also how they work, reskill, and adapt to shifting labour demands. Against this backdrop, understanding AI’s facilitative role in educational and workforce transitions is both timely and essential. Governments, institutions, and industries are investing heavily in AI-powered systems to promote personalised learning, enhance institutional efficiency, and enable real-time skills matching in employment markets. These developments raise fundamental questions: Can AI serve as a bridge between formal education and the dynamic requirements of contemporary work? How can it foster equitable transitions across socioeconomic boundaries? And what institutional and pedagogical models best support this integration? The integration of AI into education and employment demands a delicate balance: while technology promises to bridge gaps, its implementation must prioritize World Journal of Advanced Research and Reviews, 2025, 26(03), 2491-2504 2492 equitable access, ethical data use, and collaborative governance between governments, schools, and industries to ensure it serves as a tool for inclusive progress rather than exacerbating inequalities. Educational theorists such as Vygotsky and Freire have long emphasised the importance of scaffolding and dialogic learning in fostering learner autonomy and critical consciousness. The integration of AI introduces a new dimension to these pedagogies by enabling responsive, data-driven learning environments tailored to individual learners’ cognitive zones of proximal development (Vygotsky, 1978; Freire, 1970). Recent studies have expanded on these foundational insights, showing how AI systems, through adaptive feedback, predictive analytics, and intelligent tutoring, can facilitate highly personalised learning experiences (Castro et al., 2024; Gligorea et al., 2023). In essence, AI—when ethically designed and pedagogically grounded—can amplify the core principles of Vygotsky and Freire, offering new opportunities for inclusive, reflective, and deeply personalized education. Simultaneously, workforce literature has highlighted the rapid restructuring of jobs due to AI and automation. Research indicates that while AI displaces specific routine roles, it also generates new occupations requiring hybrid skill sets, such as digital literacy, emotional intelligence, and ethical reasoning (Du, 2024; Jackson and Jackson, 2024). Institutions of higher learning are thus pressured to align curricula with these emerging workforce requirements (Nam and Ha, 2023). Scholars argue that AI-driven career guidance systems can help bridge this gap, providing data-informed, personalized pathways for reskilling and job placement (Luo, 2025; Cheng and Liang, 2023). Despite these insights, existing studies often treat educational and workforce transitions as separate domains, lacking a unified framework that positions AI as an integrative force. Moreover, the literature pays insufficient attention to how AI may either exacerbate or mitigate existing digital divides, particularly in underserved or rural populations (Mariyono and Hd, 2025). To address this, scholars must build integrative frameworks that acknowledge AI’s dual role in shaping education-workforce ecosystems—while rigorously examining its effects on marginalized groups. 1.1. Research Gap While existing research has examined AI’s role in personalised learning (Castro et al., 2024) and workforce transformation (Ros and Loeung, 2025), few studies address the interstitial space—the transitional zone between formal education and employment readiness—where AI may act as a catalytic bridge. Furthermore, limited research focuses on equity-centred implementations of AI, particularly those addressing the needs of marginalised learners or digitally excluded communities. The fragmented nature of current discourse necessitates a holistic analysis of how AI enables educational-to-workforce transitions across diverse learning contexts, socioeconomic strata, and institutional settings. 1.2. Research Aim This review aims to provide a comprehensive review and critical analysis of how artificial intelligence facilitates educational and workforce transitions in the digital era. Specifically, it seeks to evaluate the extent to which AI-driven technologies: • Bridge the institutional gap between formal education and labour market needs; • Support lifelong learning, reskilling, and upskilling initiatives; • Enable equitable access to digital learning and employment opportunities across socioeconomic groups. By synthesising recent developments in AI-enabled educational technology, workforce analytics, and digital transformation strategies, this research contributes to both theory and practice in educational and labour policy design. 1.3. Review Questions The following research questions guide the review: • How does AI enhance personalized and adaptive learning to improve learner readiness for employment? • In what ways does AI support career guidance, reskilling, and upskilling for dynamic labour markets? • How do AI-integrated educational models influence transitions into digitally mediated workplaces? • What challenges and ethical considerations arise in implementing AI across educational and employment sectors? • To what extent can AI mitigate or exacerbate digital inequities in education and workforce access? This review makes several novel contributions. First, it integrates discussions from educational technology and labour studies to conceptualise AI as a transitional infrastructure—a system that links formal learning with future employment World Journal of Advanced Research and Reviews, 2025, 26(03), 2491-2504 2493 in an adaptive, scalable, and personalised manner. Second, it identifies critical leverage points for equitable AI implementation, offering evidence-based insights for policymakers and institutional leaders. Finally, it proposes a future research agenda that centres inclusivity, transparency, and systemic design in AI applications for education-toemployment transitions. 2. Methodology This study employs a narrative review methodology to synthesise and critically evaluate the role of AI as a bridging mechanism between education and workforce development in the digital era. A narrative review is particularly wellsuited for exploring complex, multidisciplinary topics that intersect various fields, such as educational technology, labour studies, artificial intelligence policy, and digital equity, where empirical evidence is emerging and theoretical frameworks are still evolving (Greenhalgh et al., 2018). 2.1. Rationale for Narrative Review Unlike systematic reviews, which focus narrowly on quantifiable outcomes and predefined search criteria, narrative reviews offer the flexibility to integrate insights from diverse research traditions, accommodate conceptual diversity, and trace thematic evolutions over time (Ferrari, 2015). Given the rapidly evolving nature of AI applications in education and employment, a narrative review allows for a broader exploration of the emerging patterns, theoretical tensions, and practical challenges across studies from 2020 to 2025. Furthermore, this method enables the incorporation of foundational theoretical contributions—from constructivist pedagogy to socio-technical systems theory—alongside cutting-edge empirical findings. The resulting synthesis offers a comprehensive understanding of how AI technologies are influencing educational and labour transitions, while also critically examining the assumptions and limitations inherent in the literature. 2.2. Data Sources and Selection Criteria To ensure academic rigour, this review included only peer-reviewed journal articles, academic books, and reputable conference proceedings published between 2020 and 2025. Foundational texts predating this window (e.g., Vygotsky, Freire, Papert) were also included where relevant to frame the conceptual foundations of AI in education. Academic databases such as Scopus, Web of Science, ERIC, and Google Scholar were searched using combinations of the following keywords: • “Artificial Intelligence in Education” • “Workforce AI transition” • “Digital transformation and employment” • “Reskilling with AI” • “AI and career readiness” • “AI in lifelong learning” • “Educational equity and AI” • “Smart universities” The inclusion criteria prioritized: • Studies offering empirical evidence or theoretical models on AI in education or employment; • Research explicitly addressing transitions between learning and workforce environments; • Interdisciplinary works combining insights from education, computer science, labour economics, and sociology. Studies were excluded if they were: • Non-peer-reviewed (e.g., blog posts, editorial opinions); • Focused solely on technical algorithm development without contextual relevance to education or work; • Redundant or lacking methodological transparency. • Irrelevant studies. In total, over 70 scholarly sources were reviewed, with 28 high-quality studies selected for detailed thematic analysis and citation. World Journal of Advanced Research and Reviews, 2025, 26(03), 2491-2504 2494 2.3. Thematic Analysis Approach Following data collection, an inductive thematic analysis was conducted to identify recurring patterns and emergent themes within the literature. This approach involved: • Initial reading and coding of texts to extract core arguments, theoretical frameworks, and reported outcomes; • Grouping of studies under conceptual categories aligned with the research questions: (a) personalized learning, (b) career guidance and reskilling, (c) institutional transformation, (d) digital equity, and (e) ethical/implementation challenges; • Synthesis and interpretation of findings to draw connections across studies, assess contradictions, and identify research gaps. While the review draws on multiple disciplines, particular attention was paid to how each work contributes to understanding AI as a mediating agent in educational and workforce transitions. 2.4. Limitations of Methodology As a narrative review, this study is inherently interpretive rather than exhaustive or replicable in the systematic sense. The selection of sources and themes reflects a curated analysis shaped by the researcher's disciplinary lens and theoretical priorities. Additionally, while care was taken to include global perspectives, there is a potential for geographic bias given the predominance of English-language publications and studies from high-income countries. Future reviews may benefit from a systematic scoping methodology or meta-analytic techniques to validate and extend the findings related to education presented here. 3. Literature Review 3.1. AI and Personalised Learning: Pedagogical Innovation in the Digital Age The integration of AI in educational settings has significantly transformed traditional pedagogical approaches, particularly through the advancement of personalized and adaptive learning. Grounded in constructivist theory (Vygotsky, 1978; Papert, 1980), AI-driven personalisation aligns with the principle that learning is most effective when it is tailored to individual learners’ cognitive and developmental needs. Recent empirical studies have highlighted how AI enables adaptive feedback, real-time content adjustments, and learner profiling, thereby supporting differentiated instruction and engagement (Castro et al., 2024; Gligorea et al., 2023). Figure 1 Timeline: Evolution of AI in Education and Workforce Systems World Journal of Advanced Research and Reviews, 2025, 26(03), 2491-2504 2495 AI-powered systems such as intelligent tutoring systems (ITS), recommendation engines, and predictive analytics tools can map learning trajectories and respond to learners’ preferences and performance patterns (Isaac, 2025). These tools offer granular insights into students' behaviour and cognition, allowing educators to intervene proactively. For example, AI-enhanced LMS platforms like Moodle can now deliver personalised content and alerts to address learning gaps before they escalate (Souhaib et al., 2024). However, scholars have also raised concerns about potential limitations of algorithmic personalisation. Nopas (2025) warns that while AI facilitates learner autonomy, it may also restrict exploration through overly prescriptive learning paths. There is also concern about the opacity of AI decision-making and its implications for student agency. Thus, a balanced implementation—one that blends human pedagogical judgment with algorithmic support—is essential. The AI-powered Higher Education Model (AHEM), proposed by Isaac (2025), exemplifies such integrated design, offering a framework that incorporates adaptive learning, AI-driven support services, and predictive models for academic success. This model advances beyond earlier educational technology frameworks by embedding AI at the systemic level. In sum, AI enables the reconfiguration of learning processes, improving accessibility, inclusivity, and efficiency. Nevertheless, ethical concerns, pedagogical implications, and data governance issues require sustained critical scrutiny. Future work must prioritise transparent algorithmic design, inclusive datasets, and the development of educators' AI literacy to ensure AI serves as a tool for pedagogical enhancement rather than control. 3.2. AI in Career Readiness and Labour Transitions The global workforce is undergoing a profound transformation, driven by the accelerated integration of artificial intelligence (AI) across economic sectors. Educational institutions are increasingly expected to not only impart disciplinary knowledge but also prepare students for evolving labour markets. As automation redefines employment landscapes, AI emerges as both a disruptor and a facilitator in career readiness and workforce transition (Du, 2024; Yadav and Jaryal, 2025). Its utility extends beyond job automation to include predictive career guidance, skill-matching algorithms, and personalised reskilling pathways. The impact of AI on labour markets is well-documented in recent literature. Ros and Loeung (2025) identify taskspecific AI deployment, domain knowledge, and machine processing capabilities as key mediators in reshaping job profiles and skill requirements. Notably, while some roles are rendered obsolete, others are reconstituted, emphasising hybrid skills such as data literacy, ethical reasoning, and interpersonal communication. This reconfiguration necessitates agile, AI-informed educational responses to bridge emerging skills gaps. In higher education, AI-enhanced career guidance systems are revolutionising how institutions support students’ professional development. These systems use machine learning to match student profiles with relevant opportunities, analyse labour market trends, and simulate future employment scenarios (Cheng and Liang, 2023). Luo (2025) demonstrates that AI can significantly improve employment counselling in local universities by integrating intelligent matching, personalised pathway generation, and real-time feedback. Such tools also promote equity by offering scalable, data-driven support to students who might lack traditional access to career resources. Reskilling and upskilling have also become critical focal points, particularly in response to Industry 4.0. The World Economic Forum predicts that by 2025, half of the global workforce will require substantial reskilling (Li, 2022). AI facilitates this process through personalized learning systems that recommend training modules, monitor progress, and validate competencies using blockchain credentials. These innovations are particularly valuable for adult learners and displaced workers navigating career pivots. Nevertheless, concerns persist regarding ethical implications, including data privacy, algorithmic bias, and the risk of reinforcing socioeconomic disparities (Jackson and Jackson, 2024). While AI provides tools to bridge education-toemployment transitions, its effective deployment demands institutional transparency, stakeholder collaboration, and alignment with human-centred values. In conclusion, AI holds significant potential to strengthen the continuity between education and employment. However, realising this promise requires more than technological innovation—it necessitates strategic, inclusive, and ethically grounded integration into career development ecosystems. World Journal of Advanced Research and Reviews, 2025, 26(03), 2491-2504 2496 3.3. Institutional Transformation and Smart Education The integration of artificial intelligence (AI) in higher education institutions is catalysing a systemic shift toward what scholars and policymakers term “smart education.” This transformation transcends mere technological adoption, encompassing strategic reconfigurations of pedagogical, administrative, and infrastructural dimensions. Smart universities—those that integrate AI, data analytics, the Internet of Things (IoT), and automation into their operational framework—represent the evolution of traditional education systems into agile, digitally responsive ecosystems (George and Wooden, 2023). At the core of this transformation is the convergence of AI technologies with data-driven decision-making processes. Institutions are increasingly leveraging AI for predictive analytics, student success modelling, academic advising, and automated administrative tasks (Imran and Almusharraf, 2024). These tools not only enhance operational efficiency but also enable proactive interventions based on learner behaviour and performance. For example, the AI-powered Higher Education Model (AHEM) supports institutional decision-making by combining adaptive learning pathways with intelligent course management and student profiling (Isaac, 2025). Quality 4.0 principles, originally developed within the manufacturing sector, have been adopted in educational settings to guide innovation and enhance institutional performance. These principles advocate for AI-enhanced personalisation, process optimisation, and continuous improvement in academic delivery (Imran and Almusharraf, 2024). Smart campuses employing these frameworks report gains in student engagement, retention, and instructional effectiveness. However, institutional AI adoption is not without controversy. Critics highlight potential risks related to the commodification of education, increased surveillance, and the erosion of academic autonomy (Chang et al., 2023). Faculty resistance is often driven by concerns over deskilling and reduced pedagogical agency, while students may raise ethical concerns about data collection and algorithmic opacity. Furthermore, institutional reliance on AI-generated assessments and content delivery raises questions about the authenticity of education and the depth of learning. The acceptance of AI-driven credentials by employers remains a debated issue. George and Wooden (2023) argue that unless qualification frameworks and industry standards evolve in tandem, there is a risk of misalignment between institutional innovation and labour market legitimacy. Moreover, disparities in AI access and integration between institutions, particularly in the Global South, exacerbate digital inequalities and raise concerns about educational equity (Mariyono and Hd, 2025). In summary, institutional transformation through AI holds vast promise for educational innovation and efficiency. However, its success hinges on ethical governance, stakeholder involvement, and a commitment to inclusivity. Strategic planning must ensure that smart education serves not just institutional goals, but also learners’ human development and societal advancement. 3.4. Equity and Digital Inclusion While artificial intelligence (AI) offers transformative potential in education and workforce development, its integration also foregrounds enduring concerns about equity, access, and social justice. The deployment of AI technologies often mirrors and amplifies existing structural inequalities, particularly in under-resourced educational contexts and marginalized communities. Scholars have emphasised that equitable AI integration requires more than technical innovation; it necessitates intentional design, inclusive policy frameworks, and critical awareness of historical disparities (Freire, 1970; Mariyono and Hd, 2025). The integration of AI in education is not technologically neutral but a social process laden with ethical implications. True equity requires dismantling the assumption that "advanced technology equals progress" and instead embedding social justice principles in every stage of AI development—from data collection to policy implementation. One of the primary barriers to equitable AI deployment is the digital divide—the uneven distribution of access to digital infrastructure, devices, and connectivity. Learners in rural or low-income settings often lack the necessary hardware or stable internet required for AI-powered learning platforms (Nopas, 2025). This infrastructural deficit limits their participation in AI-enhanced education and deepens educational stratification. Furthermore, AI systems trained on biased data risk perpetuating cultural, linguistic, and socioeconomic biases, excluding non-dominant learner groups from relevant and affirming educational content (Ogunseye et al., 2025). In response to these challenges, some scholars have proposed decentralized and community-based models of AIenhanced education. For instance, the Blockchain and Artificial Intelligence Non-Formal Education System (BANFES) offers an inclusive model that combines AI personalisation with secure blockchain credentialing in informal and low- World Journal of Advanced Research and Reviews, 2025, 26(03), 2491-2504 2497 trust settings (Nazari et al., 2024). BANFES facilitates personalised learning across geographies by decentralising content delivery and enabling learner-driven credential acquisition, particularly for displaced or underserved populations. From a pedagogical standpoint, inclusive AI design must account for cultural diversity, multilingualism, and local knowledge systems. AI tools must not only accommodate but also affirm learners’ varied identities and contexts. Mariyono and Hd (2025) emphasise the importance of collaborative co-design processes, involving educators, learners, and community stakeholders in shaping AI tools that reflect pluralistic values and lived realities. Such participatory approaches counteract top-down, technocratic deployments that often ignore grassroots needs. Moreover, ethical frameworks governing AI use in education should prioritise transparency, algorithmic explainability, and mechanisms for accountability. Equity-focused capacity-building—training educators and administrators in AI literacy and critical digital pedagogy—is crucial for bridging gaps in implementation knowledge and institutional readiness (Essa, 2024). In sum, while AI holds the potential to democratise access to quality education and career development, its equitable application depends on human-centred design, infrastructural investment, and sustained policy attention. Without such considerations, AI may not bridge the gap, but widen it. 3.5. Pedagogical and Ethical Concerns The integration of artificial intelligence (AI) in education necessitates rigorous interrogation of its pedagogical alignment and ethical implications. As AI increasingly mediates learning environments—from intelligent tutoring systems to automated assessment tools—educators and researchers must consider whether these technologies genuinely enhance learning or undermine core educational values. Drawing from foundational pedagogical theorists such as Vygotsky and Freire, the concern is not merely what AI can do, but how and why it is used in instructional contexts (Vygotsky, 1978; Freire, 1970). Pedagogically, AI offers opportunities for self-regulated learning, formative feedback, and individualised support. However, scholars caution against uncritical adoption. Chang et al. (2023) argue that AI-powered systems, such as chatbots and recommendation engines, should be designed following key pedagogical principles: goal-setting, learner autonomy, formative feedback, and cognitive scaffolding. If poorly implemented, these systems risk becoming transactional tools that limit exploration and reduce learners to data profiles (Nopas, 2025). Ethical considerations further complicate the deployment of AI in education. Central issues include algorithmic bias, surveillance, data privacy, and the erosion of human oversight. AI systems trained on biased datasets can perpetuate discriminatory practices, affecting learner evaluation, access to resources, and long-term opportunities (Ogunseye et al., 2025). Instructors and institutions may unknowingly embed these inequities into their systems, particularly when using black-box AI tools with little transparency or explainability. The rise of generative AI, such as ChatGPT and similar tools, has also reshaped debates around academic honesty and authorship. Benson (2024) notes that while such tools can enhance creativity and engagement, they also raise serious concerns about plagiarism, misrepresentation, and a decline in critical thinking. Rather than banning these technologies, scholars advocate for the development of new assessment strategies that embrace AI use while maintaining academic integrity and pedagogical rigour. Moreover, faculty preparedness remains a critical concern. Many educators lack the necessary training to critically engage with AI tools or integrate them meaningfully into their pedagogy. Without professional development and institutional support, the risk of misalignment between technology and teaching goals increases (Mosly, 2024). To address these challenges, scholars recommend multi-stakeholder collaboration in the design and implementation of AI in education, ensuring alignment with ethical standards, pedagogical principles, and learner empowerment. As Freire (1970) reminds us, education must be a practice of freedom, not merely a system of efficiency. In this light, AI must be critically interrogated to ensure it enhances rather than compromises the humanistic aims of education. World Journal of Advanced Research and Reviews, 2025, 26(03), 2491-2504 2498 Figure 2 Thematic Map: AI in Education-to-Workforce Transitions 4. Discussion The findings of this literature review demonstrate that artificial intelligence (AI) is reshaping the education-toworkforce pipeline by reconfiguring learning environments, expanding reskilling opportunities, and influencing institutional transformation. However, while AI technologies offer powerful tools for personalisation, career preparation, and equity enhancement, their integration is far from frictionless. This discussion critically examines three key domains: (1) the pedagogical redefinition of learning and agency, (2) the restructuring of workforce transitions, and (3) the socio-technical challenges of implementing equitable AI. 4.1. Redefining Pedagogy and Learner Agency AI has advanced the pedagogical landscape by operationalising principles of constructivism, particularly in fostering individualised learning pathways (Vygotsky, 1978; Papert, 1980). Adaptive learning systems reflect the tenets of the zone of proximal development by offering tailored scaffolding and just-in-time feedback (Castro et al., 2024). However, these gains must be critically assessed. Nopas (2025) argues that algorithmic mediation can paradoxically constrain learner autonomy by embedding invisible boundaries on exploration and cognitive flexibility. Similarly, the reliance on opaque recommendation engines may prioritise efficiency over inquiry, raising questions about the depth and richness of AI-mediated learning. Freire's (1970) insistence on dialogical education reminds us that learning should not become a mechanistic process driven by algorithmic predictions. Instead, AI systems must be designed to enhance, not replace, reflective pedagogy. Therefore, human-centred design principles should prioritise transparency, learner choice, and the co-creation of knowledge within digital platforms (Chang et al., 2023). 4.2. Reshaping Workforce Preparation and Lifelong Learning AI is not only changing job requirements but also altering how individuals prepare for and navigate work over a lifetime. The rise of AI-driven career guidance systems and dynamic skill-matching algorithms has enabled data-informed, individualised career pathways (Luo, 2025; Cheng and Liang, 2023). These developments align with lifelong learning models advocated by Illeris (2009); wherein continuous skill renewal is central to individual adaptability and employability. World Journal of Advanced Research and Reviews, 2025, 26(03), 2491-2504 2499 Nevertheless, the uneven implementation of such tools risks deepening the divide between those with digital access and algorithmic literacy and those without. While Li (2022) underscores the necessity for large-scale reskilling by 2025, infrastructural and institutional readiness remain uneven globally. Reskilling strategies must not only be technically robust but socially inclusive, reflecting both local labour contexts and learner diversity (Ros and Loeung, 2025; Yadav and Jaryal, 2025). Figure 3 AI-Powered Learning-to-Work Pathway 4.3. Advancing Evidence-Based Applications of AI Integration As AI becomes further embedded in education and workforce ecosystems, there is a growing imperative to expand and diversify the empirical evidence that underpins its deployment. While much of the current discourse around AI in education is conceptually rich, the need for longitudinal, comparative, and context-sensitive research is increasingly recognised as essential to informing policy and practice at scale (OECD, 2023; Essa, 2024). Global initiatives are already contributing to this foundation. The OECD’s AI and the Future of Skills Project (2023), for instance, evaluates how AI is reshaping training ecosystems across member countries through the collection of crossnational data on task transformation and employment trajectory forecasting. Similarly, UNESCO (2021) has conducted regional pilots of AI-enhanced learning environments in Latin America and Sub-Saharan Africa, offering empirically grounded insights into learner engagement, performance differentials, and challenges of infrastructure and access in lower-income contexts. These studies emphasize that AI’s educational value is maximised not only by technological sophistication but by its fit within institutional capacities and learner needs. At the national level, India’s National Digital Education Architecture (NDEAR) represents a paradigmatic model of scalable AI deployment in public education. NDEAR leverages federated architecture and AI-enhanced dashboards to provide personalised insights at both the individual learner and system levels. Initial analyses indicate improved retention, targeted interventions, and enhanced administrative responsiveness (Ministry of Education, India, 2023). In Europe, the Directorate-General for Employment has piloted AI-enabled platforms that align upskilling programs with