Artificial Intelligence in Distance Education: A Systematic Review of Emerging Pedagogical, Cognitive and Institutional Dynamics

Synthesis: This systematic review of 56 peer-reviewed articles (2020–2025) on AI in distance education identifies four interrelated themes: AI-driven personalization and adaptive support, AI-mediated assessment and feedback, human–AI interaction and pedagogical transformation, and AI AI Governance, fairness, integrity and equity. Collectively the findings show that AI is not merely a tool but a participant that reorganizes relationships among learners, teachers and institutions. The evidence base remains limited by a lack of longitudinal research, with most studies being short-term or cross-sectional.

Key Findings

• AI-driven personalization, learner modeling and adaptive support (24 studies): AI generates personalized learning pathways, constructs learner models and provides adaptive support, improving alignment between instructional design and learner needs; effectiveness depends heavily on data quality and representativeness, with risks of over-automation that can undermine learner Learner Agency.
• AI-mediated assessment, feedback and learning facilitation (19 studies): AI enhances the immediacy and pedagogical value of formative feedback, supports collaborative knowledge building and deepens content mastery, while raising concerns about reliability, transparency, Privacy and learner trust.
• Human–AI interaction, learning processes and pedagogical transformation (17 studies): AI emerges as a participant in learning rather than a background tool, reshaping cognitive, emotional and social processes; instructors' roles shift as automation frees time for individualized support, and AI literacy strongly influences learners' ability to benefit.
• AI governance, fairness, integrity and equity (19 studies): AI can promote equitable access and inclusive education when designed intentionally, but may deepen inequities when access, literacy or cultural responsiveness is uneven; effective governance requires technical safeguards plus institutional policies for trust, transparency, privacy and equity.
• Evidence limitations: Most reviewed studies relied on short-term interventions or cross-sectional data, making it difficult to assess sustained effects on learning, motivation or institutional practice; longitudinal research is needed to understand how learner–AI relationships, instructor strategies and institutional policies evolve over time.

Study Design & Method

The review used a systematic synthesis approach. Initial searches were run in the Education Source and ERIC databases using keywords "artificial intelligence or ai or a.i." and "online learning or e-learning or distance learning," restricted to peer-reviewed articles in English with full text available, published 2020–2025. Of 88 initially identified articles, 27 duplicates were removed, and five more were excluded (three out-of-range literature reviews, two conference papers without detailed findings), leaving 56 unique articles across higher education, K–12, professional training and informal learning contexts. Each article was read in full with analytic notes capturing research purpose, AI functions, learning contexts, methodology, findings and implications, with particular attention to how each study conceptualized AI's role (as tutor, tool, assessor, collaborator, facilitator of self-regulation or institutional mechanism). Iterative analysis refined early categories into four analytically robust themes, with some studies assigned to more than one theme.

What this means for practice

• Instructors. Cultivate AI literacy in learners as a precondition for benefit, and expect the instructor's role to change rather than disappear: automation frees time for individualized support only if new pedagogical competencies, collaboration, and accountability frameworks are built alongside it.
• Designers. Adopt intentional, theory-informed design and check the quality and representativeness of the data behind any personalization before trusting its recommendations; guard against over-automation that erodes learner Learner Agency.
• Administrators. Build AI Governance that pairs technical safeguards with institutional policy on transparency, Privacy, Trust and equity, keeping the field's own commitments to access, flexibility, and learner empowerment central so AI serves high-quality learning rather than efficiency alone.
• Administrators. Commission longitudinal evaluation rather than short pilots, because the review's evidence base is dominated by short-term and cross-sectional studies and cannot show sustained effects on learning, motivation, or institutional practice.
• Instructors. Treat formative feedback quality as the test of an AI tool: the reviewed studies found immediacy gains alongside unresolved problems of reliability, transparency and learner trust.

Limitations

• The analysis rests on 56 articles that were preselected and supplied in full text rather than retrieved through an exhaustive search, so the authors state it does not represent the entirety of global research on AI in distance education for 2020–2025.
• Searching was confined to two databases (Education Source and ERIC) and to English-language, peer-reviewed, full-text-available articles, excluding computational education, HCI and learning-sciences venues.
• Of 88 initially identified articles, 27 duplicates and five more items (three out-of-range literature reviews and two conference papers without detailed findings) were removed, and some retained studies were assigned to more than one of the four themes.
• Most reviewed studies used short-term interventions or cross-sectional data, so the review cannot assess sustained effects on learning, motivation or institutional practice.

Citation

Corry, M. (2026). Artificial intelligence in distance education: a systematic review. Quarterly Review of Distance Education, 27(1).