Front. Educ. Frontiers in Education Front. Educ. 2504-284X Frontiers Media S.A. 10.3389/feduc.2026.1793940 Original Research Artificial intelligence as a scaffolding tool for self-directed learning in ODeL environments: an instrumental case study of Zimbabwe Open University Magwa Logic * Conceptualization Methodology Writing – original draft Institute for Open Distance Learning, College of Education, University of South Africa, Pretoria, South Africa Correspondence: Logic Magwa, magwal@unisa.ac.za 27 02 2026 2026 11 1793940 22 01 2026 11 02 2026 19 02 2026 Copyright © 2026 Magwa. 2026 Magwa https://creativecommons.org/licenses/by/4.0/ This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

Artificial Intelligence (AI) is increasingly integrated into Open and Distance eLearning (ODeL) environments, yet limited empirical research explains how AI functions as a scaffolding mechanism to support self-directed learning (SDL), particularly in developing country contexts. This qualitative instrumental case study examined how AI tools are perceived and utilized to scaffold self-directed learning at Zimbabwe Open University. Data were generated through semi-structured interviews with lecturers and a quality assurance coordinator and focus group discussions with students who had experience in using AI- supported learning tools. Data were analyzed using Braun and Clarke’s six-phase thematic analysis. Findings indicate that participants perceived AI as supporting self-directed learning through personalized learning support, instant feedback and data-driven reflection and goal setting. While the study relies on participant perceptions rather than direct observation of AI system behaviour, it provides context-sensitive insights into how AI may function as a pedagogical scaffold in resource-constrained ODeL environments. These mechanisms function as digital scaffolds within learners’ Zone of Proximal Development, enabling autonomy, sustained engagement and self-regulation in ODeL contexts. The study contributes a context-sensitive AI-enabled scaffolding framework for SDL in resource-constrained ODeL environments and highlights implications for policy, practice, and future mixed methods research.

 artificial intelligence instrumental case study ODeL environments scaffolding self-directed learning Zimbabwe Open University The author(s) declared that financial support was not received for this work and/or its publication. section-at-acceptance Higher Education Background of the study

While ODeL systems emphasize learner autonomy, many students lack the metacognitive and self-regulatory skills required to manage learning independently (Oian et al., 2025). In Zimbabwean ODeL contexts, this challenge is compounded by limited instructor interaction, delayed feedback, and infrastructural constraints (Mukwevho, 2018; Musingafi et al., 2015). Although AI technologies are increasingly adopted in higher education, their role as structured scaffolding mechanisms for SDL remains underexplored. The study therefore seeks to examine how AI is utilized as a scaffolding tool to support SDL within the Zimbabwe Open University, focusing on learner experiences, instructional practices and institutional quality processes.

Owan et al. (2025) assert that, despite the growing integration of digital tools in ODeL, there has been little attention given to how artificial intelligence can specifically be used as a scaffolding tool to support self-directed learning. Similarly, Roe and Perkins (2024) point out that, AI remains an immature field of research, yet it is likely to define educational scholarship in the years to come. Most existing studies (Gundu 2024; Nadeem et al., 2024) focus on general applications of AI in education, such as assessment automation or content recommendation, with limited emphasis on its role in promoting learner autonomy, motivation, and self-regulation. In context of Zimbabwe and similar developing countries, this area remains largely unexplored. This gap highlights the need for research to establish how AI can be effectively leveraged to support SDL in ODeL environments. Guided by this focus, the study was framed by the following research question:

In what ways do AI-supported tools function as scaffolding mechanisms for learner autonomy, feedback and reflection in an ODeL context?

 Literature review

Existing studies on AI in education predominantly focus on assessment, automation, plagiarism detection, and content recommendation systems (Boafe et al., 2025; Gundu 2024; Nadeem et al., 2024). While these approaches improve efficiency and scalability, they often neglect learner autonomy and metacognitive development. Intelligent tutoring systems offer adaptive feedback but require high infrastructure capacity and technical expertise, limiting their applicability in resource constrained ODeL contexts. Chatbot-based AI tools provide immediacy and accessibility but risk superficial learning if not pedagogically aligned. Notably, there is paucity of research examining AI explicitly as a scaffolding mechanism grounded in learning theory, particularly within African ODeL institutions. This study addresses this gap by conceptualizing AI as a dynamic scaffold operating within learners’ Zones of Proximal Development to support SDL rather than merely delivering content or automating assessment.

While existing studies highlight the growing application of AI in higher education, many focus on efficiency, automation or assessment, with limited attention to how AI supports self-directed learning as a pedagogical scaffold. Moreover, few studies critically examine AI use in African ODeL contexts, where infrastructural limitations, digital literacy gaps and reliance on learner perceptions shape how AI is experienced. This study therefore contributes by examining AI not as a technological solution but as a perceived scaffolding mechanism situated within a specific ODeL context.

 Clarification of key concepts Artificial intelligence (AI)

Artificial intelligence (AI) is defined as computer systems capable of carrying out human-like processes such as learning, adaptation, synthesis, self-correction and the use of data for complex processing tasks (Grajeda et al., 2024). Frank (2024) asserts that AI refers to the simulation of human intelligence in machines that are programmed to perform tasks that typically require human cognitive functions, such as learning, reasoning, problem solving and decision-making. In the education setting, AI has proven to be a tool with immense potential to transform the teaching and learning process. According to Ocen et al. (2025), these tools are widely used to enhance the teaching and learning process in higher institutions of learning. However, despite the unprecedented opportunities, these tools, if not used adequately, can be challenging in educational institutions. Ocen et al. (2025) present different categories of AI tools commonly used in higher institutions of learning, such as ChatGPT, Avide note, Elicit, Perplexity, Consensus, Semantic Scholar, Research Rabbit, Scholarly, Mendeley, Zotero, ChatPDF among others. Some of these tools are used as references and data management tools, while others are used to aid and improve writing.

 AI tools description

In this study, a distinction is made between institutionally supported AI systems and learner-initiated AI tools. Institutionally supported systems include learning management systems (LMS) analytics, automated assessment feedback and officially integrated digital learning platforms utilized by Zimbabwe Open University. Learner-initiated tools refer to externally available generative AI applications such as Chat GPT and Grammarly, which students independently adopted to support writing, comprehension and revision. This distinction is critical as institutionally supported AI tools reflect formal pedagogical design and quality assurance processes, whereas learner-initiated tools highlight emergent informal scaffolding practices shaped by learner agency and access. In practice, learner-initiated tools such as ChatGPT and Grammarly were predominantly used by students to support writing, comprehension and revision activities. Institutionally supported AI systems, including LMS analytics and automated feedback tools, were primarily engaged by lecturers and quality assurance personnel for monitoring student progress, instructional support and quality assurance purposes, and although students also interacted with automated feedback outputs within the LMS.

 Self-directed learning

Self-Directed Learning (SDL) refers to the ability of individuals to take control of their learning process, including setting goals, identifying resources, and evaluating their progress. The concept is synonymous with learner autonomy and reflects the idea that students bear the primary responsibility for what and how they learn. Gutierrez and Tomas (2019) assert that autonomous learning is when students are allowed the freedom to determine their behaviour if they believe that lessons are meaningful to them. Similarly, Andriani et al. (2018) point out that autonomous learning is a concept in which the learners can take charge of their own learning. This implies that various terms, such as learner independence, autonomous learning, and independent learning, are often used interchangeably with self-directed learning. Brandt (2020) asserts that learners should take the initiative with or without the help of others in diagnosing their learning needs, formulating learning goals, identifying human and material resources for learning, choosing, and implementing appropriate learning strategies, and evaluating learning outcomes. In other words, self-directed learning gives learners the freedom and autonomy to choose what, why, how, and where of their learning. In self-directed learning, teachers should empower students by giving them significant ownership over their learning, involving them in the decision-making process, and giving them responsibility, rather than constantly monitoring, directing, and supervising them (McCormack et al., 2021). This would help to increase the level of student participation in school life. In the context of a crisis, self-directed learning becomes a crucial strategy for navigating the uncertainty and challenges presented by the situation.

 Scaffolding

Scaffolding is a teaching and learning strategy in which temporary support is provided to learners to help them accomplish tasks or understand concepts that they would not be able to do independently (Taber, 2018). Oian et al. (2025) assert that scaffolding is an educational construct defined as structured guidance provided to students to foster skill acquisition within a specific context. This guidance is increasingly integrated into digital learning environments. This implies that AI technologies can support, enhance and transform the way individuals take ownership of their own learning journey.

 Statement of the problem

While artificial intelligence (AI) is increasingly being adopted in education, there is limited research on its role as a scaffold tool to support self-directed learning (SDL), particularly in Open and Distance eLearning (ODeL) environments (Owan et al., 2025; Roe and Perkins, 2024). Given this fact, the need for this study is quite apparent to explore how AI can be effectively utilized as a scaffolding tool to support SDL in ODeL environments particularly at Zimbabwe Open University.

 Theoretical framework

This study is underpinned by Vygotsky’s social constructivist theory, particularly the concept of the Zone of Proximal Development (ZPD) and is supported by Self-Directed Learning (SDL) theory by Knowles. I propose Vygotsky’s theory to emphasize the role of scaffolding that is temporary support provided to learners to help them achieve tasks they cannot accomplish alone (Taber, 2018). In the context of ODeL, AI can function as a digital scaffold by offering personalized support, adaptive feedback and guided learning pathways (Marquardson, 2024). Brandt (2020) asserts that Knowles’s Self-Directed Learning (SDL) theory highlights the learner’s autonomy and responsibility in directing their own learning, which aligns with the goals of ODeL. However, many learners require structured support to effectively manage this responsibility. AI when strategically implemented can bridge the gap between learner independence and the need for support, thereby enhancing self-directed learning capabilities. Together, these theories provide a foundation for exploring how AI tools can be designed and implemented to support learner development within the ZPD while promoting autonomy and engagement in self-directed learning.

In this study, Vygotsky’s Zone of Proximal Development and Knowles’ Self-Directed Learning Theory informed both data collection and analysis. Interview and focus group questions explored how learners experienced guided support, feedback, reflection and increasing autonomy when engaging with AI-supported tools. During analysis, themes were examined in relation to ZPD-related processes (such as perceived guided assistance and reflection), enabling theoretical concepts to inform interpretation of the empirical data.

 Methods

This study adopted a qualitative instrumental case study design to explore how artificial intelligence is perceived to support self-directed learning within an Open and Distance eLearning context (Cresswell, 2009; Magwa and Magwa, 2015). The Masvingo Regional Campus of Zimbabwe Open University was selected as an instrumental case because it represents a typical regional ODeL campus with diverse student demographics, active use of digital learning platforms and exposure to both institutionally supported and learner-initiated AI tools. The case was intended to illuminate broader issues surrounding AI-supported self-directed learning in resource-constrained ODeL environments rather than to generate statistically generalizable findings.

Participants were purposively selected based on their direct experience with AI- supported learning or teaching. The sample comprised fourteen students drawn from multiple faculties, seven lecturers with more than 5 years of ODeL teaching experience and one quality assurance coordinator. Student participants varied in age, programme of study and digital competence, reflecting the diversity typical of ODeL learners. Lecturer participants were selected to ensure representation across faculties and familiarity with digital and AI-supported instructional practice.

To gather thick, rich and detailed descriptions to answer the research questions, semi-structured interviews that lasted between 45 to 60 min with seven lecturers and one quality assurance coordinator and focus group discussions for a duration of 70 to 90 min with 14 students were conducted. Examples of guiding questions included: “How do AI-supported tools assist you in managing your learning independently?” and “In what ways does AI-generated feedback influence your reflection, goal setting or learning strategies?” These prompts were designed to explore scaffolding processes related to autonomy, feedback and self-monitoring.

The generated data was transcribed into text, and a thematic content analysis, relying on identifying and analyzing the emerging themes from all the forms of data that were collected, was used. Following this logic, I was able to link the various opinions of participants on how AI can be effectively utilized as a scaffolding tool to support SDL in ODeL environments.

The study’s potential for rich and important data determined participant selection. Table 1 lists the participants’ details.

Participants’ details.

Category Number (n) Inclusion criteria Data collection participation
Students 14 Two students representative from each faculty male and female Focus group
Lecturers 7 Male and female with more than 5 years of teaching experience. Two from each faculty Semi-structured interviews
Quality assurance coordinator 1 They are deemed to have knowledge and experience gained by working with students and lecturers Semi-structured interviews
Ethical considerations

Ethics approval was obtained from the University of South Africa’s College of Education, Research Ethics Committee (Reference no 7673) after meeting privacy, confidentiality, informed consent, assent and voluntary participation requirements. Before entering research sites, permission was sought from the Ministry of Higher and Tertiary Education, Innovation, Science and Technology Development. The purpose of the study was explained to the participants. The participants were informed that they could withdraw from the study at any time without consequences. Friedriksen and Rhodes (2004) aver that informing the participants of their right to withdraw from a study is an important aspect of the research process. Member checking processes were conducted after the analysis process and contributed to the rigour and trustworthiness of the findings. The quality assurance criteria of dependability, conformability, transferability and credibility were employed.

 Data analysis

I used Braun and Clark (2006, p. 87) six phases of thematic analysis. Interview transcripts were read and re-read to align with the first step or familiarization of the data. In the second step, the data were coded. In this case, coding schemes were used to simplify the correlations between the data. In the third instance, patterns and repetitions in the data sets were recorded. In the fourth step the emerging themes were reviewed to check if they work in relation to the coded extracts. Themes were then refined to generate clear names in the fifth step and finally, the recurring patterns were collated into themes. I independently coded the data and developed the themes.

Initial coding focused on participants’ descriptions of AI use, perceived support, feedback experiences and learning autonomy. Codes such as “guided assistance,” “Instant correction,” “progress tracking,” and “independent learning” were clustered into broader themes. Themes were reviewed and refined to ensure coherence and alignment with the research question and theoretical framework. To enhance trustworthiness, emerging themes were compared across data sources and reflective memo-writing was used to minimize researcher bias.

 Results

Results are presented according to three themes:

Personalization of learning

Instant feedback and support

Data-driven reflection and goal setting

The themes identified were validated across multiple groups, enhancing analytical robustness. For example, personalization of learning was consistently reported by students, lecturers, and quality assurance personnel, indicating convergence of perspectives. This triangulation strengthens the credibility of the findings and demonstrates how AI scaffolding operates at learner instructional and institutional level.

Personalisation of learning

It was established that AI plays a key role in scaffolding self-directed learning by providing highly personalized experiences that support learners in progressing at their own pace and according to their unique needs. Excerpts from lecturers, a quality assurance coordinator and students are presented below to support the above finding.

In support of the above finding, a lecturer elaborated during an interview session as follows:

AI tools such as ChatGPT and adaptive e-learning platforms have made learning more personalized by tailoring content to the learner’s needs. For example, learners can request explanations at different difficulty levels, generate study summaries, or receive topic recommendations based on their progress. This ensures that each learner studies at his/her own pace and focuses on areas needing improvement. (L1).

Another lecturer added

From my observation, AI tools have made it easier to cater to individual learner differences in our ODeL programmes. Participants perceived that AI-supported systems track their progress and provide learning support that appears adaptive to their needs. This personalization allows students to take charge of their learning journey while we as lecturers can focus more on guiding and providing academic support, where it is most needed. (L4).

The quality assurance coordinator point of view was that

AI has introduced a new dimension to quality delivery. It allows us to monitor how individual students engage with course materials and to see learning patterns that help us assess whether the content meets diverse learner needs. This level of personalization was previously difficult to achieve in large ODeL classes.

This excerpt strengthens the institutional viewpoint that AI-driven personalization not only benefits learners individually but also enhances monitoring, equity, and quality assurance in the ODeL framework. When responding to the same question during focus group discussion, one of the students expressed her sentiments as follows:

The AI platforms helps me learn at my own pace. When I do not understand a topic, it gives me extra materials & quizzes until I get it right. That makes me feel more confident in my studies. (S2).

These accounts illustrate how AI functions as a scaffold within learners’ Zones of Proximal Development by offering guided support tailored to individual competence levels. Through differentiated explanations and adaptive content suggestions, learners gradually assume greater autonomy in managing their studies, reflecting key principles of self-directed learning.

 Instant feedback and support

The data reveals that AI scaffolds self-directed learning by providing instant feedback and on-demand support, helping learners build understanding, maintain momentum, and work independently without constant instructor input. During an interview session, a lecturer responded to this topic as follows:

AI-powered writing assistants like Grammarly and ChatGPT offer immediate feedback on assignments, grammar, and structure. This instant response helps learners identify mistakes and correct them without waiting for a tutor. Similarly, chatbots integrated into learning platforms assist students anytime, enhancing self-reliance and continuous learning. (L6).

Another lecturer elaborated

One of the biggest advantages of integrating AI into our teaching is the provision of instant feedback. Students no longer have to wait for manual marking or email responses, the system immediately shows them where they went wrong and suggest areas for improvement. This keeps the learning process continuous and dynamic. (L3).

These excerpts emphasize that lecturers view AI-driven instant feedback and support as a valuable scaffolding mechanism that promotes continuous learning, self-correction, and reduced instructional bottlenecks in the Zimbabwe Open University’s ODeL setting. A quality assurance coordinator elaborated on this point of view as follows:

AI-generated instant feedback ensures that students receive consistent and timely responses, which is critical for maintaining quality in self-directed learning. It bridges the gap between learner queries and instructor availability, ensuring that no student is left behind due to delayed feedback.

This excerpt highlights that from a quality assurance perspective, AI’s capacity to deliver instant, consistent, and accessible feedback strengthens teaching quality, learner autonomy, and institutional accountability-all essential for sustaining effective self-directed learning in ODeL environments. During a focus group discussion, a student responded to the same topic as follows:

Instant feedback motivates me to keep learning. It feels like I have a personal assistant that supports me anytime, even late at night when lecturers are not available. (S5).

This excerpt illustrates how learners perceive AI’s instant feedback and on-demand support as critical scaffolding tools that enhance understanding, sustain learning momentum and foster independence in ODeL environment. The immediacy of AI-generated feedback represents a form of mediated instructional dialogue, reducing dependency on delayed tutor responses while strengthening learners’ capacity for self-correction and independent problem solving.

 Data-driven reflection and goal setting

Findings highlight that AI facilitates self-directed learning by offering data-driven insights and scaffolding that support learners in reflecting on their progress, setting meaningful goals, and adjusting their learning strategies with greater precision. In support of this view, one of the lecturers had this to say:

AI tools provide performance analytics that help learners reflect on their progress. For instance, learning management systems (LMS) at ZOU track assignment submissions, time spent on tasks, and quiz results. This data enables learners to set realistic goals, monitor improvement, and plan study schedules effectively based on their strengths and weaknesses. (L2).

Another lecturer elaborated

In the past, students rarely reflected on their learning until exam time. Now with AI tools providing continuous progress data, they can monitor their growth week by week and make necessary adjustments. This has improved learner engagement & accountability. (L7).

These excerpts underscore lecturers’ recognition that AI-generated learning, analytics and progress tracking foster reflection, goal setting and adaptive learning strategies-key elements in scaffolding self-directed learning within ODeL framework.

AI-generated learning analytics provide valuable evidence for both students and the instruction. They allow learners to reflect on their performance while helping us evaluate whether our instructional design truly supports self-directed learning.

The implication of the above excerpt is that AI’s data driven insights serve a dual purpose-empowering learners to engage in reflection and goal setting, while also supporting institutional monitoring and quality assurance in promoting effective self-directed learning across ZOU’s ODeL programmes. During a focus group discussion session, when responding to the same question, a student narrated his experiences in the following manner:

I like how the system gives me progress reports after every unit. Seeing my scores and study hours helps me set clear goals for the next week. It’s like having a mirror for my learning progress. (S12).

This excerpt shows how learners perceive AI as a tool that helps them track progress, reflect meaningfully, and set realistic learning goals in the ODeL context. By visualizing performance data and progress indicators, AI tools support metacognitive regulation, enabling learners to monitor, evaluate and strategically adjust their learning practices-central components of self-directed learning theory.

 Discussion

This study aimed to elicit how artificial intelligence can serve as a scaffolding tool to enhance self-directed learning. Artificial Intelligence was examined through participants lived experiences of interacting with AI-enabled tools such as ChatGPT, adaptive learning platforms, learning management systems analytics, and automated feedback. Rather than testing AI systems experimentally, the study explored how these tools functioned pedagogically as scaffolds that guided learning, supported autonomy and enabled self-monitoring within ODeL environments. The findings are then explored in reference to the literature and theory.

According to Frank (2024), the integration of AI into education has ushered in transformative changes, particularly in the realm of self-directed learning. AI technologies, including machine learning algorithms, natural language processing and data analytics, are being harnessed to innovate and enhance self-directed learning Frank (2024). This implies that by providing content and activities that resonate with students on a personal level, AI make self-directed learning more relevant and enjoyable. This study affirms that personalization of learning via AI scaffolding has transformative potential for enhancing learner autonomy, motivation and engagement in ODeL settings (Merino-Campos, 2025). Al Nabhani et al. (2025) assert that success of self-directed learning depends on equitable access, digital readiness, and the strategic integration of human and artificial intelligence. The contributions of this study lie not only in demonstrating what AI personalization can achieve, but also in critically identifying the conditions under which it can thrive in resource-constrained environments. The study also highlights the pivotal role of artificial intelligence in enabling the personalization of learning within the Zimbabwe Open University’s ODeL environment, enhancing self-directed learning by tailoring content, feedback and pacing to individual learner needs. AI-powered scaffolding tools were found to support learner autonomy, motivation and engagement, particularly for ZOU’s diverse student population with varying levels of digital access and academic preparedness. Atif Wafik et al. (2025) point out that self-directed learning bridges learning gaps, promoted self-regulation, and improved learner outcomes, though challenges such as infrastructure limitations, digital literacy and unequal access remain. The findings contribute to the growing body of knowledge on how AI can be contextually applied in developing countries to support personalized, learner centered ODeL models, while emplacing the continued need for human tutor support and institutional readiness. Through the lens of Vygotsky’s Social Constructivist Theory, personalization of learning reflects the idea of scaffolding within the learner’s Zone of Proximal Development ZPD, where AI tools adapt instruction and support to the learner’s current level of competence (Marquardson, 2024). This creates a dynamic, interactive learning environment that bridges the gap between what learners can do independently and what they can achieve with assistance. From the perspective of Self-Directed Learning Theory, personalization enhances learner autonomy by allowing individuals to take charge of their learning path, set personal goals and access resources that align with their needs and preferences (Brandt, 2020). Thus, AI becomes a facilitator of both guided and independent learning process.

Merino-Campos (2025) assert AI technologies can significantly optimize self-directed learning by tailoring instant feedback and support to individual learner needs. Similarly, Atif Wafik et al. (2025) point out that AI through intelligent tutoring systems can assess student interactions to provide immediate feedback or to infer the student’s knowledge level and serve as an appropriate learning resource. This implies that instant feedback and support emerge as a critical enabler of self-directed learning in ODeL environments. Learners often operate in isolation, with limited real-time access to tutors, making timely support essential for maintaining motivation and academic progress. The study found that AI-driven systems significantly enhanced the immediacy and relevance of feedback by providing real-time responses tailored to individual learner inputs and performance (Al Nabhani et al., 2025). This immediate support enabled students to identify errors, clarify misunderstanding and reinforce learning concepts without waiting for tutor feedback, which is often delayed in traditional ODeL settings. Instant feedback also promoted learner autonomy by encouraging self-correction and reflection, key components of self-regulated learning. Moreover AI-enabled support systems-such as intelligent chatbots, automated quizzes, and adaptive learning platforms-served as always available digital tutors, reducing learners’ cognitive load and frustration (Atif Wafik et al., 2025). However, the study also noted that while instant feedback improved the learning experience, its effectiveness depended on the clarity, accuracy and contextual relevance of AI-generated responses. This underscores the importance of aligning AI tools with curriculum goals and ensuring they are culturally and linguistically appropriate. Overall, instant feedback and support through AI scaffolding proved vital in enhancing learner confidence, engagement, and persistence in the ZOU, ODeL context. In relation to Vygotsky’s Social Constructivist Theory, instant feedback and support provided by AI function as a form of mediated interaction, helping learners construct understanding through guided responses like teacher-learner dialogue. The immediate nature of feedback enables continuous scaffolding, ensuring that misconceptions are addressed promptly. Under the Self-Directed Learning Theory, instant feedback strengthens learner’s ability to monitor and evaluate their own progress, fostering self-regulation and responsibility for learning outcomes (Brandt, 2020). This real time support encourages reflection, adjustment of strategies and greater learner confidence in navigating their studies independently.

Recent empirical studies from fragile and conflict-affected higher education contexts further illuminate the opportunities and tensions associated with AI-supported self-directed learning. Studies conducted in Palestinian higher education institutions (Hamamra et al., 2025a,b) reveal that while AI enhances learner autonomy and instructional efficiency, educators experience significant technostress, ethical uncertainty and institutional readiness challenges. These findings resonate with the Zimbabwe Open University context, where AI scaffolding supports learner independence but remains dependent on infrastructural capacity, staff preparedness and policy coherence. Similarly, Khlaif et al. (2025) highlight that educators’ technostress emerges from blurred boundaries between human pedagogy and algorithmic decision-making, underscoring the need for balanced AI-human mediation. This aligns with the present study’s findings that AI functions most effectively as a scaffold when it complements-rather than replaces-lecturer guidance within learners Zones of Proximal Development. Furthermore, Alhur et al. (2025) caution against AI dependency among educators, noting a paradox whereby increased reliance on generative AI may undermine pedagogical agency. This study extends that discourse by demonstrating that in resource-constrained ODeL environments, intentional pedagogical alignment and institutional oversight are critical to sustaining AI-enabled self-directed learning without eroding academic judgement or learner responsibility.

According to Atif Wafik et al. (2025), educational technologies such as machine learning-based learning management systems, track and analyze learner behaviour, enabling instructors to make data-driven choices that improve their teaching and support students as effectively as possible. This implies that data-driven reflection and goal setting emerged as a transformative element in supporting self-directed learning within the Zimbabwe Open University’s ODeL framework. AI tools enabled the continuous collection and analysis of learner performance data, which is then presented to students in form of personalized dashboards, progress trackers and performance analytics. These insights empowered learners to reflect meaningfully on their learning habits, identify strengths and weaknesses and set realistic academic goals based on evidence rather than guesswork (Al Nabhani et al., 2025). This reflective practice-enhanced by data visualization and timely prompts-fostered metacognitive awareness and strengthened learners’ ability to plan, monitor & adjust their study strategies accordingly. For many ODeL learners, particularly those studying in isolation, this type of structured self-assessment provided a sense of direction and control over their learning journey. The study found that when learners regularly engaged with their own data, they became more motivated, focused, and accountable Al Nabhani et al. (2025). However, the effectiveness of this approach depended on learners’ digital literacy and their ability to interpret and act on the feedback. Therefore, the integration of AI-driven reflection tools must be accompanied by guidance and support to ensure all learners can benefit. Overall, data driven reflection & goal setting played a key role in fostering learner autonomy, accountability, and sustained engagement in the ODeL environment. From a Vygotskian standpoint, data-driven reflection and goal setting align with the concept of mediation where AI tools act as cognitive scaffolds that support learners in interpreting performance data and internalizing effective learning strategies (Marquardson, 2024). The process promotes metacognitive growth as learners move from assisted to independent functioning. In the context of Self-Directed Learning Theory, the use of data empowers learners to engage in self-assessment, reflect on achievements and gaps and set informed, realistic goals. This data-informed reflection reinforces self-management skills and supports a cycle of continuous improvement in open & distance e-learning environments.

The novelty of this study lies in repositioning AI from a productivity or automation tool to a pedagogical scaffold that mediates learning within ODeL contexts. Unlike prior models that emphasize content delivery, this framework foregrounds learner agency, reflective practice and adaptive support, offering a theoretical grounded contribution to ODeL scholarship in developing-country setting.

Despite the perceived benefits of AI—supported learning, participants also implicitly highlighted risks related to over-reliance on AI, uneven digital literacy and access constraints. Because many findings are based on learner and lecturer perceptions rather than direct observation of AI functionality, there is a risk of attributing pedagogical effectiveness to AI tools without sufficient evidence of learning depth. These constraints underscore the need for careful instruction design, AI literacy training and continued human mediation in ODeL environments.

Proposed AI-enabled scaffolding framework

This study proposes a conceptual AI-enabled scaffolding framework for SDL in ODeL environments comprising three interrelated components, (1) personalization of learning, (2) instant feedback and (3) data-driven reflection and goal setting. The framework assumes learner access to basic digital tools and instructional support structures. AI functions as a cognitive scaffold by adapting learning support within the learner’s Zone of Proximal Development. As a conceptual qualitative model, it does not rely on computational convergence criteria or algorithmic parameters but instead illustrates pedagogical relationships between AI functionalities and SDL processes.

 Conclusion

This study demonstrated that AI could function effectively as a scaffolding tool to support self-directed learning in ODeL environments through personalization, instant feedback and data-driven reflection. Using a qualitative case study approach, the research provided contextual insights into how AI supports learner autonomy within Zimbabwe Open University. While findings are limited by the single-site design and qualitative scope, they offer a foundation for future mixed method and large-scale studies examining learning outcomes, ethical implications and infrastructure readiness for AI integration in ODeL.

Recommendations

Based on the findings, ODeL institutions should:

Priorities AI literacy training for both students and lecturers.

Strengthen instructional design support for AI -integrated course and develop institutional policies that guide ethical, equitable and pedagogical sound use of AI in self-directed learning.

 Limitations

This study has several limitations. First findings rely primarily on participant perceptions of AI use rather than direct observation or system-level analysis of AI functionality. Second, the study was conducted at a single regional campus, limiting broader generalizability. Third, variations in participants’ digital literacy and access to AI tools may have influenced their experiences. These limitations suggest the need for future mixed method & multi-site studies incorporating learning analysis and observational data.

 Data availability statement

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

 Ethics statement

The studies involving humans were approved by the University of South Africa’s College of Education, Research Ethics Committee (Reference no 7673) after meeting privacy, confidentiality, informed consent, assent and voluntary participation requirements. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

 Author contributions

LM: Conceptualization, Methodology, Writing – original draft.

 Acknowledgments

The author(s) thank Zimbabwe’s Ministry of Higher and Tertiary Education, Innovation, Science and Technology Development for permitting me to undertake my fieldwork at Zimbabwe Open University, Masvingo Regional Campus and all the participants for sharing their experiences with me.

 Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

 Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

 Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

 Author disclaimer

The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy or position of any affiliated agency of the authors.

 References Al NabhaniF. HamzahM. B. AbuhassnaH. (2025). The role of artificial intelligence in personalizing educational content: enhancing the learning experience and developing the teacher’s role in an integrated educational environment. Contemp. Educ. Technol. 17:ep573. doi: 10.30935/cedtech/16089 AlhurA. A. KhlaifZ. N. HamamraB. Hussein (2025). Paradox of AI in higher education: qualitative inquiry into AI dependency among educators in Palestine. JMIR Med. Educ. 11:e74947. doi: 10.2196/74947 AndrianiP. F. PadmadewiN. N. BudasiI. G. (2018). Promoting autonomous learning in English through the implementation of content and language integrated learning (GLIL) in science and Maths subjects. SHS Web of Conference Atif WafikH. M. PrinceS. PromiseA. RahmanM. JahanJ. MallikS. K. (2025). Enhancing personalized learning through artificial intelligence in modern education systems. Am. J. Environ. Econ. 4:5441. doi: 10.54536/ajee.v4i1.5441 BoafeF. AnimP. A. ArthurE. (2025). Effective ways of assessment in the AI era. Empowering education: navigating the landscape of meaningful student engagement, Lancaster university education conference BrandtW. C. (2020). Measuring success skills: a review of the literature on self-directed learning. Dover: National Centre for Improvement of Educational Assessment. BraunV. ClarkeV. (2006). Using thematic analysis in psychology. Qual. Res. Psychol. 3, 77–101. doi: 10.1191/1478088706qp0630a CresswellJ. W. (2009). Research design: Qualitative, quantitative and mixed methods approaches. Los Angeles, CA: Sage. FrankE. (2024). The influence of artificial intelligence on education: enhancing personalized learning experiences. Comput. Sci. Eng. 11–25. FriedriksenK. RhodesJ. (2004). The role of teacher relationship in the lives of students. New Dir. Youth Dev. 103, 45–54. doi: 10.1002/yd.90, 15595327 GrajedaA. BurgosJ. CordovaP. SanjinesA. (2024). Assessing student-perceived impact of using artificial intelligence tools: construction of a synthetic index of application in higher education. Cogent Educ. 11:2287917. doi: 10.1080/233186x.2023.2287917 GunduT. (2024). Strategies for e-assessments in the era of generative artificial intelligence. Electron. J. E-Learn. 22, 40–50. doi: 10.34190/eje/.22.7.3474 GutierrezM. TomasJ. M. (2019). The role of perceived autonomy support in predicting university students’ academic success mediated by academic self-efficacy and school engagement. Educ. Psychol. 39, 729–748. doi: 10.1080/01443410.2019.1566519 HamamraB. KhlaifZ. N. MayalehA. (2025a). Between promise and precarity: Palestinian university educators’ experiences with AI integration in higher education. Interact. Learn. Environ., 1–61. doi: 10.1080/10494820.2025.2547264 HamamraB. KhlaifZ. N. MayalehA. BakerA. A. (2025b). A phenomenological examination of educators’ experience with AI integration in Palestinian higher education. Cogent Educ. 12:2526435. doi: 10.1080/2331186X.2025.2526435 KhlaifZ. N. HamamraB. BensalemE. MitwallyM. A. SanmugamM. (2025). Factors influencing educators’ technostress while using generative AI: a qualitative study. Technol. Knowl. Learn., 1–26. doi: 10.1007/s1007/s10758-025-09885-z MagwaS. MagwaW. (2015). A guide to conducting research. a student handbook. Singapore: Strategic Book Publishing and Rights Company. MarquardsonJ. (2024). Embracing artificial intelligence to improve self-directed learning: a cybersecurity classroom study. Inf. Syst. Educ. J. 22, 4–13. doi: 10.62273/WZBY3952 McCormackO. FlahertyJ. O. LiddyM. (2021). Students views on their participation in publicly managed second level schools in Ireland: the importance of student-teacher relationships. Educ. Stud. 47, 422–437. doi: 10.1080/03055698.2019.1706041 Merino-CamposC. (2025). The impact of artificial intelligence on personalized learning in higher education: a systematic review. Trends High. Educ. 4:17. doi: 10.3390/higheredu4020017 MukwevhoM. H. (2018). Time management challenges on students’ academic performance: a case study of a rural University in Limpopo Province, South Africa. Journal of Gender, Information and Development in Africa (JGIDA). 31. doi: 10.31920/2050-4284/2018/s1n1.12 MusingafiM. C. C. MapurangaB. ChiwanzaK. ZebronS. (2015). Challenges of open and distance learning (ODL) students: experiences from students of the Zimbabwe Open University. J. Educ. Pract. 6, 64–70. NadeemM. FaragW. A. HelahM. (2024), Rethinking assessment methodologies in the era of artificial intelligence: expanding beyond ChatGPT’s scope. Conference: Mediterranean Smart Cities Conference OcenS. ElasuJ. AarakitS. M. OlupotC. (2025). Artificial intelligence in higher education institutions: review of innovations, opportunities and challenges. Front. Educ. 10:1530247. doi: 10.3389/feduc.2025.1530247 OianK. LiS. T. RakovioM. LiX. GuanR. MolenaarI. . (2025). Towards reliable generative AI-driven scaffolding: reducing hallucinations and enhancing quality in self-regulated learning support. Comput. Edu. 240:105448. doi: 10.1016/jcompedu.2025.105448 OwanV. J. ChukwuC. O. AgamaV. U. OwanT. J. OgarJ. O. EtortiI. J. (2025). Acceptance and use of artificial intelligence for self-directed research learning among postgraduate students in Nigerian public universities. Discov. Educ. 4:329. doi: 10.1007/s44217-025-00770-6 RoeJ. PerkinsM. (2024). Generative AI in self-directed learning: a scoping review. Comput. Sci. 1:07677. doi: 10.48550/arXiv.2411.07677 TaberK. S. (2018). “Scaffolding learning: principles for effective teaching and the design of classroom resources” in Effective teaching and learning: Perspectives strategies and implementation. ed. AbendM. (New York, NY: Nova Science Publishers).

Edited by: Zuheir N. Khlaif, An-Najah National University, Palestine

Reviewed by: Gülten Genç, İnönü University, Türkiye

Bilal Tawfiq Hamamra, An-Najah National University, Palestine