Artificial Intelligence in Advancing Quality Education: A Systematic Literature Review

1. Introduction

The rapid advancement of digital technologies is reshaping global education systems, creating both opportunities and challenges for achieving SDG 4 [36,38]. Among these innovations, Artificial Intelligence (AI) has emerged as a transformative tool with the potential to enhance access, improve learning quality, and support lifelong education [26,27]. By enabling personalised learning, automated feedback, and scalable content delivery, AI offers new pathways to address persistent educational inequalities [5]. However, its integration raises critical concerns regarding equity, ethics, and governance [19,34].

Education is widely recognised as a foundational driver of sustainable development, social equity, and economic growth [3]. In 2015, the United Nations adopted the SDG 4 as a central pillar aimed at ensuring inclusive and equitable quality education and promoting lifelong learning opportunities for all. Multiple authors [25,27,36], alluded that despite global commitment, progress towards SDG 4 remains uneven. While the study of [18], observed persistent challenges such continued lack of gender parity in education particularly in low and middle income countries. More challenges identified are inadequate infrastructure [38]; teacher shortage most acute in sub–Saharan Africa [7], learning poverty [20], and digital divides [2,23], continue to undermine educational outcomes. However, study of [19] emphasized that AI is transforming the foundations of academic integrity into higher education, not encouraging individual misconduct.

In addition to structural challenges, the rapid digitalisation of education systems has introduced new complexities related to technological readiness and policy alignment. The integration of digital technologies in education has been uneven across regions, with high-income countries benefiting from advanced infrastructure, while low- and middle-income countries continue to face constraints in connectivity, affordability, and digital skills ([36,38]. This disparity has significant implications for achieving SDG 4, as the digital divide increasingly shapes access to quality education. In this context, emerging technologies such as AI offer both opportunities and risks, necessitating a critical examination of their role in promoting equitable and sustainable educational outcomes.

AI, which creates content such as text, images, and audio based on user prompts, enables personalised learning environments by adapting curricula and improving feedback systems. Further it’s regarded as large language models and generative content tools, that can produce human-like text, adaptive feedback, instructional materials, and data-driven insights at scale. AI has emerged as a potentially transformative force in education. These capabilities offer new possibilities for addressing long-standing educational challenges, particularly those related to access, personalisation, and quality of instruction. Prior research highlights that AI can support personalised learning pathways, intelligent tutoring systems, automated assessment, and multilingual content generation, thereby aligning closely with several SDG 4 targets.

Existing literature reveals a concentration of research in higher education and technologically advanced regions, with comparatively limited empirical evidence on AI’s impact in primary education, early childhood learning, and the Global South. To address this gap, the present study conducts a systematic review of the literature published between 2016 and 2025. Although several reviews and bibliometric studies have examined AI or AI in education broadly, there remains a lack of systematic, SDG-focused synthesis that explicitly maps AI contributions to SDG 4. Current studies often explore pedagogical effectiveness or technological adoption in isolation, without adequately situating findings within the SDG4 framework or assessing implications for inclusive and sustainable educational development. The objective of this study is to critically evaluate the role of AI particularly large language models and content-generation systems in advancing equitable access to education, enhancing the quality of learning outcomes, and promoting inclusive, lifelong learning within diverse and resource-constrained contexts.

The transformation of educational delivery from traditional classroom-based instruction to remote, virtual, and blended learning environments has accelerated significantly in recent years [32]. Further, the study of [22] alluded that this shift has been further intensified by the emergence of AI, which is increasingly being integrated into teaching and learning processes. Existing studies highlight that AI has the potential to enhance learning efficiency, engagement, and accessibility through personalised content generation, adaptive feedback, and intelligent tutoring systems (9, 17, 33]. However, while these technological advancements offer promising opportunities, their implications vary across different educational contexts, particularly in relation to access, quality, and lifelong learning.

From a theoretical perspective, the integration of AI in education can be understood through the lens of Human Capital Theory, which emphasises the role of education in enhancing productivity and economic development, and Digital Inclusion Theory, which highlights the importance of equitable access to digital resources. While AI enhances knowledge acquisition and skill development, disparities in access to technology may reinforce existing inequalities [12,39]. This duality underscores the need for a balanced approach that considers both technological innovation and social equity.

A critical analysis of the literature reveals that most studies adopt a techno-optimistic perspective, emphasising the benefits of AI while underreporting its limitations. For instance, while studies such as [5,27] highlight improved learning outcomes and engagement, fewer studies provide robust empirical evidence on long-term educational impact, particularly in developing contexts. Additionally, the predominance of higher education-focused research limits the generalisability of findings to primary and secondary education systems. This indicates a need for more context-sensitive and longitudinal research to better understand the broader implications of AI in education.

Beyond the identified themes, emerging literature highlights the growing importance of contextual adaptability in the implementation of AI within education systems. While AI tools offer scalable and flexible solutions, their effectiveness varies significantly depending on socio-economic, institutional, and technological contexts. Studies such as [12,39] emphasise that digital inequality is not solely a function of access to technology, but also of disparities in digital literacy, institutional support, and policy readiness. This suggests that the successful integration of AI requires a holistic ecosystem that includes infrastructure, training, and governance mechanisms.

Furthermore, recent research indicates that AI is reshaping the role of educators from knowledge transmitters to facilitators of learning. [22] argues that the integration of AI in classrooms necessitates a collaborative model where human expertise and machine intelligence complement each other. This shift has implications for curriculum design, assessment practices, and professional development, requiring educators to acquire new competencies in AI literacy and digital pedagogy.

In addition, there is growing recognition of the need for ethical and responsible AI use in education. [34] highlights the importance of establishing frameworks that ensure transparency, accountability, and fairness in AI-driven educational systems. Without such safeguards, there is a risk that AI may reinforce existing biases and inequalities, particularly in algorithmic decision-making processes. Therefore, literature increasingly calls for a balanced approach that integrates technological innovation with ethical considerations and inclusive design principles.

The literature demonstrates that AI can play a significant role in improving equitable access to education [17,39], particularly in resource-constrained and underserved contexts. [27] emphasise the importance of integrating innovative digital technologies and sustainable educational practices to advance SDG 4. While their empirical study, based on survey data from 197 participants, reveals that both educators and students perceive AI as a valuable tool for enhancing teaching effectiveness and expanding learning opportunities. The findings further indicate that AI supports the development of critical competencies, such as decision-making and entrepreneurial thinking, thereby contributing to more inclusive and transformative educational environments.

Despite these benefits, the literature also suggests that access remains uneven, particularly in regions with limited digital infrastructure [12]. While AI has the capacity to democratise education through scalable and flexible learning platforms [41], its effectiveness is contingent upon the availability of technological resources and digital literacy. This highlights the need for policy interventions and institutional support to ensure that AI-driven educational innovations do not exacerbate existing inequalities.

A substantial body of research focuses on the role of AI in enhancing the quality of learning outcomes and promoting inclusive educational practices. [28] examines the application of AI in architecture education, demonstrating its potential to transform teaching and learning through improved design visualisation, creative problem-solving, and enhanced student engagement. Using a mixed-methods approach, the study identifies significant improvements in students’ critical thinking abilities and overall learning experiences.

More broadly, AI is shown to facilitate personalised learning pathways, automated assessment, and real-time feedback, which contribute to improved academic performance and learner satisfaction. However, the literature also raises important concerns regarding academic integrity, over-reliance on AI-generated outputs, and the need to redesign assessment frameworks. These challenges underscore the importance of balancing technological innovation with ethical considerations and pedagogical integrity to ensure sustainable improvements in educational quality.

The role of AI in promoting lifelong learning has also gained increasing attention, particularly in the context of rapidly evolving knowledge economies. [14] provides insights into the application of artificial intelligence in surgical education, highlighting its potential to enhance learner engagement, standardise training, and expand global access to high-quality educational resources. Their study, based on expert consultations and literature review, demonstrates that AI can support continuous professional development and reduce variability in training outcomes.

However, the integration of AI into lifelong learning systems is not without challenges. Issues related to academic integrity, evolving faculty roles, assessment redesign, and ethical considerations remain significant [4]. Furthermore, disparities in access to digital technologies continue to limit the scalability of these innovations in resource-constrained environments. Consequently, while AI offers substantial opportunities for advancing lifelong learning, its successful implementation requires careful consideration of governance frameworks, institutional readiness, and ethical deployment.

2. Materials and Methods

This systematic review aimed to synthesise the existing literature on the role of AI in advancing sustainable development goal 4: quality education. The transparent, repeatable, and iterative nature of the review process led to the utilisation of a Systematic Literature Review methodology (SLR). This method provides an impartial basis for excluding research that is not directly relevant, addressing the issue of subjectivity found in traditional reviews. The widespread availability of electronic databases has facilitated timely and systematic research.

In response, this systematic review aims to critically examine the role of AI in advancing SDG 4: Quality Education. By synthesising empirical and conceptual literature, the study seeks to (i) identify key AI applications aligned with SDG 4 targets, (ii) assess reported educational benefits and risks, and (iii) highlight research gaps and policy implications for sustainable and ethical integration of AI in education systems. Through this contribution, the review supports a more coherent understanding of how AI can be leveraged responsibly to accelerate progress towards inclusive, equitable, and high-quality education.

The review employed a comprehensive search strategy on Scopus databases, to locate pertinent peer reviewed journal articles, conference proceedings, and other scholarly sources. The search terms encompassed keywords related to the “Artificial Intelligence” AND “Sustainable Development Goal 4” AND “Quality Education”. This systematic literature review followed the PRISMA guidelines [40]. Relevant data from the selected studies were extracted and synthesised, (see Table 1).

To further enhance the rigor of the review, a structured coding framework was applied during the data extraction process. Each selected study was systematically analysed based on predefined criteria, including research design, methodological rigor, data quality, theoretical grounding, and relevance to SDG 4 themes. This approach enabled the identification of patterns and relationships across studies, facilitating a more robust thematic synthesis.

Additionally, efforts were made to minimise bias in the selection and analysis of studies. The use of clearly defined inclusion and exclusion criteria ensured consistency in the screening process, while the application of quality appraisal measures helped to filter out low-quality or irrelevant studies. Although the reliance on a single database (Scopus) may limit the comprehensiveness of the review, it provides a reliable and high-quality source of peer-reviewed literature. The synthesis process involved both descriptive and interpretive analysis, allowing for the identification of key trends, gaps, and emerging issues within the literature. This dual approach strengthens the validity of the findings and supports the development of a comprehensive understanding of the role of AI in advancing SDG 4.

The review employed a systematic search strategy using the Scopus database to identify relevant literature on AI and SDG 4. Table 2 illustrates the studies selected based on predefined inclusion criteria, focusing on peer-reviewed journal articles published between 2016 and 2025 that examine the role of AI in advancing quality education. Exclusion criteria eliminated conference papers, books, non-English publications where translation was not feasible, and studies published prior to 2016. The selected articles were then subjected to a structured data extraction and synthesis process, with emphasis on key themes related to access, learning quality, and lifelong learning.

The identification and screening process presented in Table 3 demonstrates a systematic and rigorous approach to selecting relevant studies. The initial search using Scopus, based on the keywords “Artificial Intelligence,” “Sustainable Development Goal 4,” and “Quality Education,” yielded 40 articles. During the screening phase, one book was excluded to maintain consistency with the inclusion criteria focused on peer-reviewed journal articles. A further refinement process led to the removal of two untraceable studies, ensuring the reliability and accessibility of the dataset. Consequently, the final sample comprised 37 relevant articles, reflecting a relatively low exclusion rate and indicating a well-targeted search strategy. This process enhances the credibility and transparency of the review, aligning with PRISMA principles and ensuring that only high-quality and relevant studies were included in the final analysis.

The PRISMA framework ensured transparency in the study selection process by systematically documenting each stage of identification, screening, eligibility, and inclusion. The search was conducted using Scopus due to its comprehensive coverage of peer-reviewed literature. The use of Boolean operators ensured precision in capturing relevant studies. Furthermore, quality appraisal criteria were applied to evaluate methodological rigor, data validity, and theoretical grounding, thereby enhancing the reliability of the findings. The inclusion of both conceptual and empirical studies enabled a holistic understanding of the research landscape, although it introduced variability in methodological approaches.

3. Results and Discussion

The thematic analysis of the Scopus indexed studies reveals a clear concentration of research on enhancing the quality of learning outcomes, with a strong emphasis on personalised learning, assessment innovation, and student engagement. A substantial body of literature also addresses inclusive and lifelong learning, particularly in relation to AI literacy, continuous skills development, and institutional readiness for digital transformation. In contrast, equitable access to education is comparatively less represented, although critical contributions highlight the role of AI in supporting accessibility, multilingual learning, and inclusion for underserved populations.

A deeper examination of the findings reveals important interconnections between the three thematic areas of access, quality, and lifelong learning. While these themes are presented separately for analytical clarity, they are inherently interdependent in practice. For example, improvements in learning quality through personalised and adaptive learning systems are contingent upon equitable access to digital technologies. Similarly, the promotion of lifelong learning is closely linked to both access and quality, as continuous learning opportunities depend on the availability of inclusive and high-quality educational resources. Moreover, the findings suggest that institutional readiness plays a critical role in determining the success of AI implementation. Institutions with well-developed digital infrastructures, supportive leadership, and clear policy frameworks are better positioned to leverage AI effectively. In contrast, institutions in resource-constrained environments face significant challenges in adopting and integrating these technologies, which may exacerbate existing inequalities.

Another key insight is the evolving role of learners in AI-enabled educational environments. The shift towards self-directed and personalised learning requires learners to develop higher levels of autonomy, digital literacy, and critical thinking skills. This transformation aligns with the broader objectives of SDG 4, which emphasise not only access to education but also the development of relevant skills for sustainable development. However, the increasing reliance on AI also raises concerns about the potential erosion of fundamental learning processes. Over-dependence on AI-generated content may limit opportunities for deep learning and critical reflection, particularly if not carefully managed. Therefore, educators and policymakers must strike a balance between leveraging the benefits of AI and preserving the core principles of effective teaching and learning.

Theme 1: Equitable access to education

Table 4 illustrates that AI enhances equitable access to education by addressing barriers related to disability, language, and affordability. [15,30] show how AI supports learners with disabilities, while [8] suggest that culturally adaptive AI studies emphasise multilingual and inclusive learning environments. [21] highlights the role of open educational resources, and Sotelo et al. (2025) demonstrate cost-effective solutions for resource-constrained settings. However, [26] reveal persistent digital divides in African HEIs, limiting adoption and access. Additional inclusion-focused studies confirm that infrastructure and institutional readiness remain key constraints. Overall, while AI expands access, its impact depends on equitable infrastructure, policy support, and digital inclusion strategies.

Theme 2: Enhancing the quality of learning outcomes

Table 4 illustrates that most studies focus on how AI improves the quality of learning through personalised and adaptive learning. [27,28] shows enhanced creativity and engagement, while [5,30] highlights improved teaching effectiveness and student experience. [37] further demonstrate gains in knowledge construction and critical thinking. In assessment, [10,11] shows improved feedback and performance outcomes. However, [20] raise concerns about academic integrity and assessment reliability. Overall, AI enhances learning quality but requires ethical safeguards and redesigned assessment frameworks.

Theme 3 Promoting Inclusive and Lifelong Learning

Table 4 illustrates that AI also supports inclusive and lifelong learning by enabling continuous, self-directed education. [6] highlight its role in skill development and learner autonomy, while [24,29] emphasise its broader systemic impact. [1], demonstrate how emerging technologies enhance innovation and entrepreneurship, and Sandhu et al. (2024) stress the importance of AI literacy. [33] introduce a human-centered perspective through Ubuntu principles. Meanwhile, [13,32] highlight governance and ethical considerations. Overall, AI enables lifelong learning but depends on institutional readiness, policy alignment, and responsible implementation.

The results indicate a clear imbalance in the distribution of research across SDG 4 themes, with a dominant focus on improving learning outcomes and comparatively limited attention to equitable access. This imbalance reflects broader systemic inequalities in global education systems, where technological innovations are often developed and implemented in resource-rich environments. Furthermore, the findings suggest that while AI enhances pedagogical innovation, its effectiveness is mediated by contextual factors such as institutional readiness, educator competencies, and policy frameworks. These findings align with Digital Inclusion Theory, which emphasises that access to technology alone is insufficient without the necessary skills and support systems.

4. Conclusions

This systematic review provides a critical synthesis of how AI, particularly LLMs, contributes to advancing SDG 4 through three core dimensions: equitable access, quality of learning outcomes, and inclusive lifelong learning. The findings reveal that AI holds transformative potential; however, its impact remains uneven and context dependent.

Firstly, with respect to equitable access, the evidence indicates that AI can significantly reduce barriers to education through multilingual content generation, assistive technologies, and scalable digital platforms. However, this potential is constrained by persistent digital inequalities, particularly in low- and middle-income countries where infrastructure, connectivity, and digital literacy remain limited.

Secondly, in relation to enhancing the quality of learning outcomes, the review highlights that AI improves pedagogical practices through personalised learning pathways, automated assessment, and intelligent tutoring systems. Empirical evidence shows that AI fosters student engagement, critical thinking, and creativity. However, concerns regarding academic integrity, over-reliance on AI-generated outputs, and the need to redesign assessment strategies remain significant.

Thirdly, regarding inclusive and lifelong learning, the findings demonstrate that AI supports continuous learning through flexible, self-directed, and skills-based education models. Nonetheless, uneven adoption across regions and institutions suggests that its contribution remains emerging.

This study concludes that AI has significant potential to transform education systems by improving equitable access, enhancing the quality of learning outcomes, and supporting lifelong learning. The findings indicate that AI enables personalised and adaptive learning, automated assessment, and scalable educational delivery, thereby aligning closely with the objectives of SDG 4 [5,25,27]. Furthermore, AI facilitates continuous learning and skills development in the context of Education 4.0, supporting more flexible and learner-centred educational models [6,24]. However, the benefits of AI remain unevenly distributed due to persistent digital divides, infrastructural limitations, and disparities in digital literacy, particularly in low- and middle-income countries [2,26]. In addition, ethical concerns related to academic integrity, algorithmic bias, and data privacy pose significant challenges to its responsible adoption [19,34]. As such, AI should be viewed not as a standalone solution, but as an enabler within a broader ecosystem that requires robust policy frameworks, institutional governance, and sustained investment in digital capacity building.

This study has several limitations. First, it is limited to Scopus-indexed publications, which may exclude relevant studies from other databases and grey literature. Second, the focus on English-language publications may result in the omission of important global perspectives, particularly from non-English-speaking regions. Third, the inclusion of both conceptual and empirical studies introduces variability in methodological rigor, which may influence the consistency of findings. Fourth, there is a noticeable bias toward higher education contexts, with limited empirical evidence from primary education, early childhood learning, and informal education systems. Finally, given the rapid evolution of AI technologies, the findings of this review may quickly become outdated, necessitating continuous updates to capture emerging developments [29,35].

Despite these limitations, this study makes several important contributions. It provides a systematic, SDG 4-aligned synthesis of AI applications in education, offering a structured thematic framework that links AI to equitable access, learning quality, and lifelong learning. By integrating insights from Human Capital Theory and Digital Inclusion Theory, the study advances theoretical understanding of how digital technologies influence educational outcomes and inequalities. Furthermore, it offers policy-relevant recommendations for governments, educators, and institutions, emphasising the need for inclusive design, ethical governance, and strategic investment in digital infrastructure. Finally, the study identifies critical research gaps, particularly the need for more empirical studies in underrepresented contexts and longitudinal research to assess the long-term impact of AI on educational systems and sustainable development.

The findings of this study have important implications for policymakers and educational institutions. First, there is a need to invest in digital infrastructure to ensure equitable access to AI technologies. Second, capacity-building initiatives should be implemented to enhance digital literacy among educators and learners. Third, ethical guidelines and governance frameworks must be established to address concerns related to data privacy, algorithmic bias, and academic integrity. Finally, policymakers should promote inclusive innovation by supporting research and implementation in underrepresented contexts, particularly in low-income regions.