Nurlan Mammadov,
Doctor of Philosophy in Pedagogy
Department of Pedagogy
Baku Slavic University
Azerbaijan, Baku
ORCID: 0000-0002-9673-0862
Keywords: artificial intelligence, learning psychology, motivation, cognitive load, self-regulated learning, AI-supported education, adaptive learning, metacognition, learner autonomy, human-AI interaction
Abstract
The increasing integration of artificial intelligence into education is transforming not only instructional technologies but also the psychological mechanisms underlying learning. AI-supported educational environments provide opportunities for adaptive instruction, individualized feedback, intelligent tutoring, continuous assessment, and personalized learning trajectories. At the same time, the growing involvement of artificial intelligence in cognitive and educational activities raises important questions concerning learner motivation, cognitive load, self-regulation, autonomy, critical thinking, and psychological dependence on technological assistance.
This article examines motivation, cognitive load, and self-regulated learning as interconnected psychological dimensions of artificial intelligence-supported education. Particular attention is given to the ways in which AI-based systems may influence learners’ engagement, perceived competence, goal setting, attention, decision-making, and regulation of their own learning activities. The study approaches artificial intelligence not as an autonomous substitute for the teacher or learner but as a pedagogical instrument whose psychological effects depend on the purposes, conditions, and methods of its use. The analysis also considers the potential benefits and risks of personalized learning, generative AI, automated feedback, and human–AI interaction in educational settings. It is argued that the effectiveness of AI-supported learning should be evaluated not only in terms of technological functionality and academic performance but also through its capacity to preserve learner agency, encourage intellectual effort, and support sustainable motivation and self-regulation. In this context, a psychologically informed pedagogical approach is necessary to ensure that artificial intelligence enhances rather than replaces the cognitive and metacognitive processes essential to meaningful learning.
Main part. Artificial intelligence-supported education should be considered not only as a technological development but also as a transformation of the psychological conditions under which learning takes place. Traditional digital technologies mainly provide access to information and learning resources, whereas contemporary AI systems can interact with learners, generate content, adapt tasks, provide immediate feedback, and modify learning activities according to individual performance. As a result, the learner is increasingly engaged in an environment where part of the cognitive and organizational work may be shared with an intelligent system. This transformation makes motivation, cognitive load, self-regulation, and learner autonomy particularly important psychological dimensions of AI-supported learning.
Motivation determines the direction, intensity, and persistence of learning behavior. In educational environments supported by artificial intelligence, motivation may be influenced by personalization, immediate feedback, perceived competence, and the possibility of progressing at an individually appropriate pace. Adaptive systems can identify areas in which a learner experiences difficulty and modify the sequence or complexity of tasks accordingly. Such responsiveness may reduce repeated experiences of failure and help learners perceive academic goals as achievable. From the perspective of self-determination theory, sustainable motivation is closely related to the satisfaction of the psychological needs for autonomy, competence, and relatedness [Ryan & Deci, 2020].
AI-supported learning can strengthen perceived competence when feedback is timely and tasks correspond to the learner’s current level of knowledge. It may also support autonomy by allowing students to make choices concerning learning pace, resources, or pathways. However, personalization should not be equated automatically with motivation. If the system makes most decisions for the learner, determines every subsequent activity, or provides solutions before sufficient intellectual effort has occurred, technological personalization may paradoxically reduce learner autonomy.
This issue is particularly relevant to generative artificial intelligence. A student can now obtain explanations, examples, summaries, draft texts, and solutions within seconds. From a motivational perspective, such accessibility can be beneficial when AI functions as a tutor that provides clarification and encourages further inquiry. Conversely, if the learner repeatedly delegates difficult cognitive tasks to the system, motivation may gradually shift from mastering a problem toward obtaining an immediate answer. The pedagogical value of AI therefore depends partly on whether it encourages intellectual participation or facilitates avoidance of intellectual effort.
The transformative character of this relationship has also been discussed in Azerbaijani educational research. Aliyev and Mammadov consider AI as a factor capable of reshaping educational philosophy, pedagogical practice, and human–machine cooperation. Their analysis emphasizes personalized learning trajectories, adaptive instruction, real-time analytics, and the changing distribution of functions between teachers, learners, and intelligent systems [Aliyev & Mammadov, 2026]. From a psychological perspective, this transformation requires particular attention to how technological assistance influences students’ responsibility for their own learning.
Another important psychological dimension of AI-supported education concerns cognitive load. Working memory has limited capacity, and learning becomes less effective when instructional demands exceed the cognitive resources available to the learner. Cognitive load theory distinguishes between demands associated with the complexity of the learning material and unnecessary demands created by the way information or activities are organized [Sweller et al., 2019]. Artificial intelligence can potentially reduce unnecessary cognitive load. For example, an AI-supported system may divide complex material into manageable stages, provide explanations when difficulties arise, identify errors, or recommend relevant resources. Instead of spending substantial cognitive effort searching through large amounts of information, students may receive content corresponding more closely to their immediate learning needs. In this respect, AI can function as a form of cognitive support. Nevertheless, reducing cognitive load is pedagogically useful only when the system removes unnecessary difficulty rather than the intellectual effort required for learning itself. Productive learning often requires learners to recall information, compare alternatives, formulate arguments, solve problems, and correct errors. If AI performs these processes on behalf of students, the immediate task may become easier while opportunities for developing independent cognitive abilities may decrease. This distinction is especially important in generative AI environments. The production of a coherent answer by an AI system does not necessarily mean that the learner has understood the underlying concepts. A student may obtain a high-quality response without engaging sufficiently in analysis, retrieval, synthesis, or evaluation. Thus, educational effectiveness should not be assessed solely by the speed or quality of the final product. The cognitive processes through which that product is achieved must also be considered.
Self-regulated learning refers to learners’ active management of their goals, strategies, behavior, motivation, and progress. Self-regulated learners plan their activities, select appropriate learning strategies, monitor understanding, evaluate results, and modify their behavior when necessary [Zimmerman, 2002]. These processes become particularly significant in AI-supported environments because students are increasingly required to decide when, why, and how technological assistance should be used. Artificial intelligence can provide useful support for self-regulation. Adaptive platforms may display learning progress, identify recurring difficulties, suggest revision, provide formative feedback, and assist students in planning subsequent activities. Such functions can make learning processes more visible and help learners recognize gaps between current performance and intended goals. Research on AI in education has identified personalization, intelligent tutoring, assessment, and learner profiling among the major areas in which artificial intelligence can support educational processes [Zawacki-Richter et al., 2019]. At the same time, externally provided regulation should not replace self-regulation. If an intelligent system continuously selects tasks, identifies errors, proposes corrections, and determines what should be learned next, students may have fewer opportunities to develop their own planning and monitoring strategies. Therefore, an important pedagogical principle is the gradual transfer of responsibility from technological support to the learner. AI should assist students in learning how to regulate their activity rather than permanently regulating that activity for them. Metacognition is closely connected with this process. Learners need to assess not only whether an AI-generated response appears convincing but also whether it is accurate, relevant, sufficiently supported, and consistent with the requirements of the task. Generative systems may produce plausible but inaccurate information, which makes critical evaluation an essential component of AI literacy. Consequently, interaction with artificial intelligence can become a valuable metacognitive activity when students are required to question, verify, compare, and improve generated responses rather than accept them automatically.
The psychological consequences of AI-supported education are also associated with the changing relationship between human agency and technological assistance. Educational technologies have historically supported learners through calculators, search engines, digital libraries, and learning management systems. Generative AI differs from many earlier tools because it can participate directly in activities traditionally associated with human cognitive production, including explanation, interpretation, writing, problem solving, and idea generation. Aliyev and Mammadov emphasize that contemporary AI creates new forms of human–machine cooperation in education and therefore requires reconsideration of the respective roles of the learner, teacher, and technology [Aliyev & Mammadov, 2026]. Psychologically, the central issue is not whether students use AI but whether they remain active agents in the learning process. The learner should retain responsibility for defining the problem, evaluating information, making decisions, and reflecting on the result. This principle is particularly important for maintaining critical and creative thinking. If AI-generated content is used as material for analysis, comparison, revision, or debate, it can stimulate higher-order thinking. For example, students may be asked to identify weaknesses in an AI-generated argument, compare several responses, verify claims against academic sources, or improve an initial solution. In such cases, AI becomes an object and partner of cognitive activity rather than a substitute for it.
The integration of artificial intelligence does not diminish the psychological importance of the teacher. On the contrary, as automated systems assume some functions related to information delivery, feedback, and routine assessment, the teacher’s role in motivation, emotional support, pedagogical judgment, and the development of learner autonomy becomes increasingly significant. AI can identify patterns in performance, but the educational meaning of those patterns requires professional interpretation within the broader context of the learner’s development. Teachers also play an essential role in establishing psychologically appropriate boundaries for AI use. Students need to understand which activities may appropriately involve technological assistance and which require independent performance. These boundaries should be determined by learning objectives rather than by a simple distinction between permitted and prohibited technologies. If the purpose of a task is to develop argumentation, for example, submitting an AI-generated argument without critical engagement undermines the educational objective. If the purpose is to evaluate arguments, however, an AI-generated text may provide useful material for analysis. Thus, AI literacy should include psychological and metacognitive competencies alongside technical skills. Students need to recognize when technological assistance supports learning, when it creates distraction or unnecessary dependence, and when independent cognitive effort is necessary. Teachers, in turn, require sufficient pedagogical competence to design AI-supported tasks that maintain these distinctions.
The effective integration of artificial intelligence into education ultimately requires a balance between technological assistance and human cognitive activity. Personalized recommendations, adaptive tasks, immediate feedback, and intelligent tutoring can contribute to motivation and reduce unnecessary cognitive demands. Yet the same mechanisms may become counterproductive if they remove productive difficulty, weaken learner autonomy, or encourage excessive dependence on external assistance. For this reason, the psychological effectiveness of AI-supported learning should be evaluated through several interconnected indicators: whether learners remain motivated to understand rather than merely complete tasks; whether cognitive support facilitates rather than replaces meaningful mental effort; whether students progressively develop self-regulation; and whether interaction with AI strengthens critical evaluation and independent decision-making. These criteria shift attention from what artificial intelligence is technically capable of doing toward what learners psychologically and educationally gain from using it. The transformative potential of artificial intelligence in education therefore lies not simply in automation but in the possibility of reorganizing learning around more responsive forms of support and human–machine interaction [Aliyev & Mammadov, 2026]. From a psychological and pedagogical perspective, however, this potential can be realized only when technology is designed and used to strengthen human agency. Artificial intelligence should expand the learner’s opportunities to understand, reflect, create, and regulate learning rather than become a substitute for these processes.
Conclusion
Artificial intelligence is creating new psychological and pedagogical conditions for learning by changing how students access information, receive feedback, solve problems, and regulate their educational activities. Its influence on learning cannot be evaluated solely in terms of technological efficiency, since the educational value of AI also depends on its effects on motivation, cognitive load, autonomy, and self-regulation.
The analysis shows that AI-supported learning can strengthen motivation and provide individualized cognitive support when it is purposefully integrated into the educational process. At the same time, excessive reliance on automated assistance may reduce independent intellectual effort and limit opportunities for developing self-regulatory and metacognitive skills. Therefore, artificial intelligence should support rather than replace the cognitive processes through which meaningful learning occurs.
In conclusion, an effective model of AI-supported education requires a balanced relationship between technological assistance and learner agency. The central objective should be to use artificial intelligence in ways that encourage active thinking, sustained motivation, critical evaluation, and responsibility for one’s own learning. Such an approach allows the possibilities of AI to be integrated with the psychological principles of learning and the pedagogical objectives of contemporary education.
References
- Aliyev, P., & Mammadov, N. (2026). Süni intellektin təhsil fəlsəfəsini, pedaqoji təcrübəni və insan-maşın əməkdaşlığını yenidən formalaşdıran transformativ potensialı [The transformative potential of artificial intelligence in reshaping educational philosophy, pedagogical practice, and human–machine collaboration]. Azərbaycan Məktəbi, 1(714), 43–51. https://doi.org/10.30546/32898065.2026.01.101
- Ryan, R. M., & Deci, E. L. (2020). Intrinsic and extrinsic motivation from a self-determination theory perspective: Definitions, theory, practices, and future directions. Contemporary Educational Psychology, 61, 101860. https://doi.org/10.1016/j.cedpsych.2020.101860
- Sweller, J., van Merriënboer, J. J. G., & Paas, F. (2019). Cognitive architecture and instructional design: 20 years later. Educational Psychology Review, 31, 261–292. https://doi.org/10.1007/s10648-019-09465-5
- Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education – Where are the educators? International Journal of Educational Technology in Higher Education, 16, Article 39. https://doi.org/10.1186/s41239-019-0171-0
- Zimmerman, B. J. (2002). Becoming a self-regulated learner: An overview. Theory Into Practice, 41(2), 64–70. https://doi.org/10.1207/S15430421TIP4102_2