Motivation and emotion/Book/2025/AI use, cognitive load, and motivation

AI use, cognitive load, and motivation:
How does generative AI reduce cognitive effort, and what are the motivational consequences?

Overview

Scenario

David has an essay due at the end of the week and feels overwhelmed. He struggles to start, as his mind is occupied with recalling concepts, planning the structure, and worrying about the deadline. Turning to a generative AI tool, David quickly receives a structured outline with key points and examples. This reduces his cognitive load, allowing him to focus on refining arguments and adding personal insights. As he gains a sense of control over the assignment and experiences increased competence, he begins to enjoy the task and views it as an opportunity to develop his skills.

Cognitive load theory suggests that working memory has limited capacity and can only hold so much information at once. To manage this, tasks can be offloaded to external tools, preserving mental effort. Generative AI is emerging as a powerful tool for cognitive offloading, providing immediate, personalised support for routine and complex tasks. This can make tasks more manageable and enhance feelings of competence and autonomy. However, over-reliance on AI may reduce intrinsic motivation by limiting task engagement, independent problem-solving, and confidence in completing tasks unassisted. It is therefore important to examine how AI reduces cognitive load and what the motivational consequences might be. Although AI can be applied across various contexts to reduce cognitive load and influence motivation, this chapter primarily focuses on learners and educational settings, reflecting the main contexts in which research has been conducted.

Focus questions

What is cognitive load theory?
How does generative AI reduce cognitive load?
What is self-determination theory?
How does reducing cognitive load with generative AI affect motivation?

Generative artificial intelligence

Figure 2. Generative AI tools

OpenAI's ChatGPT

Gemini by Google

Generative artificial intelligence (GenAI) refers to technologies that use machine learning to create new content, such as text, images, audio, and video (Martin et al., 2025). AI systems, which were once limited to routine and repetitive tasks, can now work alongside humans on more complex and cognitively demanding problems by analysing patterns in large datasets and generating outputs in response to human input (Wu et al., 2025). There has been rapid uptake of GenAI around the world, transforming the nature and structure of many professional tasks and reshaping teaching and learning in educational settings (Martin et al., 2025; Wu et al., 2025). Frequently used GenAI applications include ChatGPT, Claude, Microsoft Copilot, and Gemini (See figure 2) (Martin et al., 2025).

Cognitive load and cognitive offloading

[Include an introductory paragraph before branching into sub-sections]

Cognitive load theory

Cognitive load theory explains the amount of mental effort required to process information. Human working memory has limited capacity, which constrains how much information can be processed at once. The theory highlights the load imposed on working memory by a given task and emphasises that instructional design should optimise cognitive resources to reduce overload and enhance learning efficiency (Gkintoni et al., 2025; Sweller et al., 2011). Broadly, there are three types of cognitive load:

Intrinsic load refers to the inherent difficulty of the material being learned. Some tasks are naturally more complex than others, such as solving advanced mathematical problems compared to simple arithmetic. The level of intrinsic load depends on the nature of the content and the learner’ level of knowledge, with more unfamiliar or abstract information placing greater demands on working memory. As such, intrinsic cognitive load cannot be altered except by altering what is learned or the levels of expertise of the learners (Sweller et al., 2011). In an educational context, teachers can manage intrinsic cognitive load by presenting instructional material that is appropriate to students' level of knowledge (Martin et al., 2021).

Extraneous load is caused by the way information is structured and presented (Martin et al., 2021). It occurs when unclear instructions, irrelevant details, or poorly designed materials place unnecessary demands on working memory. Such overload interferes with learning, as learners must process too many elements simultaneously (Sweller et al., 2011). Unlike intrinsic cognitive load, extraneous cognitive load should always be reduced and under no circumstances increased (Sweller et al., 2011). In an educational context, teachers can manage extraneous load by presenting instructional material sequentially, clearly, and explicitly to students (Martin et al., 2021). Cognitive load theory is primarily concerned with identifying techniques to reduce extraneous cognitive load (Sweller et al., 2010).

Germane load facilitates meaningful cognitive processing that is essential for schema formation and deep learning (Gkintoni et al., 2025). It relates to the working memory resources that the learner uses to process and understand the information (Sweller, 2010). While extraneous load should be minimised, germane load is beneficial and can be encouraged through instructional techniques like self-explanation, elaboration, and active retrieval practice (Gkintoni et al., 2025). For example, if students are prompted to question the rationale behind factual information rather than simply memorising it, they actively engage with the material, creating richer connections between schemata, improving long-term retention (Gkintoni et al., 2025).

Table 1.

Types of Cognitive Load

Intrinsic load	Inherent difficulty of the task (Can’t easily change)
Extraneous load	Unnecessary effort caused by poor design (Should be reduced)
Germane load	Meaningful cognitive processing (Should be encouraged)

Cognitive offloading

Figure 3. External tools

Calculator

Internet search engine

Cognitive offloading, as described by Risko and Gilbert (2016) involves using external tools to reduce the mental effort and cognitive demands of a task. Because working memory is limited in capacity, delegating parts of a task to external aids allows individuals to conserve cognitive resources (Risko & Gilbert, 2016). Research shows that offloading can make tasks more manageable and improve performance across domains such as memory, arithmetic, counting, and spatial reasoning (Risko & Gilbert, 2016). Calculators and the internet are examples of external tools that have transformed the way individuals process, store, and retrieve information (see Figure 3) (Gerlich, 2025). GenAI has emerged as an extension on these tools, capable of performing complex tasks, generating content, and supporting decision-making (Gerlich, 2025).

Generative AI and cognitive load

GenAI has the potential to reduce cognitive effort across multiple dimensions of cognitive load. While Intrinsic load, the inherent complexity of a task, cannot be directly reduced by AI, AI tools can make tasks more manageable by slowing the pace of explanations, providing examples, simplifying concepts, or presenting information in multiple formats. For more advanced students, AI can offer additional challenges to prevent boredom (Hussain et al., 2025). AI can also minimise extraneous load caused by irrelevant or distracting information by filtering out unnecessary information, highlighting important content, organising tasks into smaller, logically ordered blocks, and improving clarity and pacing (Gerlich, 2025; Hussain et al., 2025). Beyond managing difficulty and distractions, AI can also enhance germane load by promoting meaningful learning through interactive and engaging features such as quizzes, flashcards, and simulations, which encourage active recall, self-assessment, and schema development (Hussain et al., 2025).

Table 2.

How AI Helps to Manages Three Types of Cognitive Load

Intrinsic load	AI can help simplify content to the learner’s level
Extraneous load	AI can remove unnecessary information or clarify instructions
Germane load	AI can structure material and highlight meaningful patterns

Reducing cognitive load using GenAI has clear benefits: tasks can become more approachable, manageable, and efficient, allowing cognitive resources to be directed to higher-order thinking. While utilisation of AI may be highly beneficial for managing cognitive load, there are also concerns around the consequences excessive reliance on offloading may also have. Reducing the need for engagement in deep, reflective thinking may undermine the intrinsic motivation that drives learning and growth (Fan et al., 2025; Wu et al., 2025).

Motivational theories

Motivation can be defined as the driving force that influences a person’s choices, actions and persistence (Galindo-Domínguez et al., 2025). A large body of theories have emerged to understand individual motivation, such as expectancy-value theory, attribution theory, social cognitive theory, and goal-orientation theory. Self-determination theory (SDT) is one of the most significant frameworks for understanding human motivation (Ryan & Deci, 2000).

Basic psychological needs theory

One mini-theory of SDT is the basic psychological needs theory, which proposes that human well-being and optimal functioning depends on the universal needs for autonomy, competence, and relatedness. The theory argues that all three needs are essential and need to be supported or there will be distinct functional costs (Ryan & Deci, 2000).

Competence

Competence refers to the need for an individual to develop their skills and experience a sense of mastery (Galindo-Domínguez et al., 2025). AI has the potential to create competence through customised learning, providing tailored learning methods and adapting to accommodate a person’s strengths and weaknesses (Jose et al., 2025; Wu, 2023). This relates to the concept of self-efficacy, which reflects an individual’s belief in their ability to succeed in specific tasks. By offering personalised support and ongoing feedback, AI can strengthen self-efficacy and, in turn, enhance learners’ sense of competence, confidence, and motivation (Wu, 2023). While AI can support competence through personalised learning, other research warns that over-reliance on AI tools may diminish learners’ sense of self-efficacy, and undermine competence (Fan et al., 2025; Jose et al., 2025).

Autonomy

Autonomy refers to the need to feel a sense of choice in one’s actions and to pursue activities that align with personal values and interests (Galindo-Domínguez et al., 2025). Task engagement increases when individuals perceive that they have control and are the primary agents of their actions (Wu et al., 2025). By providing personalised learning paths, AI tools may support students to engage in activities that align with their interests, allowing them to take ownership of their learning. However, reliance on AI or a lack of ownership over AI-generated content, may weaken an individual’s sense of autonomy (Fan et al., 2025; Jose et al., 2025).

Relatedness

Relatedness refers to the need for a sense of social connectedness and belonging (Galindo-Domínguez et al., 2025). The immediate responsiveness of AI tools and personalised outputs can make individuals feel connected during otherwise solitary tasks, and as AI becomes more human-like, it may further support this need. However, if users turn to AI instead of seeking guidance from peers and teachers, relatedness may be inhibited (Jose et al., 2025). If engaging with AI tools does provide a sense of relatedness, it may increase the dependence on AI tools as tasks without AI assistance may feel less enjoyable (Wu et al., 2025).

Intrinsic and extrinsic motivation

Intrinsic motivation refers to the desire to engage in a task out of enjoyment, curiosity, or a sense of fulfilment. People who are intrinsically motivated complete a task because it feels rewarding or meaningful. Extrinsic motivation, on the other hand, occurs when people engage in a task for the outcome, such as rewards, punishments, or a response from others. Figure 4 shows two children playing sports: one plays for enjoyment, representing intrinsic motivation, while the other plays to earn a reward, illustrating extrinsic motivation. Intrinsic motivation is the preferred motivator of behaviour as it drives high-quality engagement, supporting sustained effort and deeper satisfaction.

According to SDT, intrinsic motivation arises from fulfilment of the three basic psychological needs (Ryan & Deci, 2000). AI tools can impact these motivational factors either positively or negatively. If AI supports these needs, it fosters intrinsic motivation; if it undermines them, motivation becomes more extrinsic (Wu et al., 2025).

Motivational consequences of AI-based offloading

[Include an introductory paragraph before branching into sub-sections]

Positive effects

High cognitive load can overwhelm students, undermining their sense of competence and motivation. The following research suggests that AI tools can reduce this load by personalising learning, enhancing focus, and supporting autonomy, creating conditions that foster intrinsic motivation and deeper engagement.

Hussain et al. (2025) discuss that experiencing high cognitive load adds pressure and stress, which reduces students' interests in learning. It can also affect emotional control, with learners having a tendency of giving up easily when they face academic challenges. From an SDT perspective, when the complexity of a task leads to cognitive overload, learners are less likely to feel capable and have their sense of autonomy and competence supported, which in turn undermines their intrinsic motivation to engage.

Building on this reasoning, Hussain et al. (2025) sought to evaluate whether AI could make learning easier, more interesting, and more meaningful by reducing cognitive load and personalising responses to learners’ needs. They conducted a study with a sample of 250 university students in Punjab, using a quantitative correlational design to examine associations between AI use, cognitive load, focus, and memory. Their findings revealed a moderate, statistically significant negative correlation between cognitive load and student focus (r = –0.457, p < .001), suggesting that when AI reduces cognitive load, students' engagement with tasks improves. When cognitive load is reduced and feedback is personalised, students may engage more fully with the material, supporting their intrinsic motivation, as described in SDT.

Emerging research has expanded on the value of AI tools in supporting highly personalised learning. Gkintoni et al. (2025) conducted a systematic review demonstrating how AI-driven adaptive learning systems, informed by real-time neurophysiological data such as EEG, can optimise cognitive load. By monitoring mental effort, these systems can adjust learning pathways, pacing, difficulty, and feedback to align with each student’s individual needs. The review notes that tools such as portable EEG headsets and brain–computer interfaces could one day be used in classrooms to provide AI with real-time information about an individual student’s cognitive state. Although widespread classroom use of EEG technology may not be practical, the findings are promising: AI can interpret neurophysiological data to understand the relationship between brain activity, cognitive load, and emotions, and adapt learning accordingly. Such insights are valuable for training AI models to recognise cognitive overload and adapt output to support student engagement and feelings of competence.

This research and other studies (e.g., Byers, 2024; Chen & Chang, 2024) support the idea that AI can enhance autonomy and competence by reducing cognitive load. It is also theorised that the human-like nature of interactions with AI, and its ability to provide immediate, highly personalised support may also enhance relatedness. However, Chiu et al. (2024) note that this area has not been extensively studied, and it remains unclear how AI, as a form of digital support, satisfies the need for relatedness.

Most research suggesting that AI tools can foster intrinsic motivation focuses on short-term effects, with limited evidence of long-term benefits. However, Hussain et al. (2025) provides findings with implications for longer-term outcomes. They found that AI tools enhance deep comprehension and learning by providing opportunities for metacognitive practice, including self-assessment and reflection. Their regression analyses showed a strong positive correlation between AI use and students’ ability to retain information (β = 0.582). If AI improves knowledge retention, learners may approach future tasks with greater perceived autonomy and competence, fostering intrinsic motivation.

Key points

AI tools can reduce mental effort, making tasks easier and more meaningful.
Personalised AI feedback supports deeper engagement and learning.
Deeper task engagement supports autonomy and competence.
AI may support relatedness through human-like interaction.
By improving knowledge retention, AI could have long-term benefits on autonomy and competence.

Negative effects

While AI can reduce cognitive load and momentarily increase engagement, several studies suggest these benefits may not translate into sustained intrinsic motivation. Fan et al. (2025) conducted an experimental study with 117 university students to explore the effect of AI support on learners’ intrinsic motivation, relative to support from a human expert, a checklist tool, or no support (control). While no significant differences in intrinsic motivation were found across the four groups, descriptive statistics indicated that the control group reported the lowest interest and enjoyment and the highest pressure and tension. Notably, the checklist group reported the highest levels of interest, enjoyment and perceived competence. These findings suggest that receiving any form of support, including AI, may reduce cognitive load by easing pressure and tension and improve short-term engagement, however AI does not appear to enhance perceived competence any more than other supports.

Moreover, Fan et al. (2025) raised concerns that by lowering cognitive effort, AI may limit opportunities for deeper learning and long-term motivational development. Excessive reliance on AI could prevent learners from experiencing the disfluency or cognitive difficulty necessary to engage deeper metacognitive processes. This contrasts with Hussain et al. (2025), who suggested that AI can increase germane load and support sustained intrinsic motivation. These differing findings indicate that AI’s impact on germane load, and by extension, on autonomy and competence, may depend on whether it is designed to support learners’ cognitive processing or to replace it.

Providing further evidence that AI may limit opportunities for meaningful mental effort, Chen et al. (2025) conducted an experimental study with 160 college students and found that heavy reliance on AI led to measurable declines in cognitive engagement, raising concerns about potential long-term effects on human cognition. Similarly, Gerlich (2025), in a mixed-method study of 669 participants across diverse ages, educational levels, and professions, reported that increased AI use was associated with lower critical thinking skills, with cognitive offloading acting as a mediating factor.

As AI tools reduce engagement in critical thinking, users may become accustomed to the ease and convenience of AI-provided solutions (Gerlich, 2025). Jose et al. (2025) emphasise that engagement encompasses both the quality and extent of mental effort. Over-reliance on AI not only reduces this engagement but also impairs the development of independent problem-solving abilities. When learners are no longer challenged to learn and solve problems independently, they may struggle to maintain the sense of satisfaction and competence that reinforces intrinsic motivation (Wu, 2025). If the thinking is done by AI, students may lose not only cognitive engagement but also the intrinsic motivation to learn and solve problems on their own (Jose et al., 2025). This reduction in independent effort can undermine learners’ autonomy, as they may no longer see themselves as the primary agents of their learning, which can also diminish their sense of competence by limiting opportunities to overcome challenges and experience achievement. The need for relatedness can also be undermined when learners rely heavily on AI. If students engage less in collaborative problem-solving, discussion, or peer feedback because AI provides easy solutions, they may feel isolated or disconnected from the learning community. If the needs of autonomy, competence, and relatedness are not supported, individuals are less likely to sustain intrinsic motivation, reducing both engagement and the long-term development of independent learning skills.

Key points

AI may reduce cognitive load, but not necessarily better than other supports.
Over-reliance on AI may limit opportunities for deeper learning and metacognitive engagement.
Heavy AI use can reduce independent problem-solving, lowering satisfaction and perceived competence.
Learners may experience reduced autonomy, seeing themselves as less responsible for their learning.
AI can undermine relatedness by decreasing peer interaction.

Conclusion

AI tools reduce cognitive effort by offloading working memory demands, minimising extraneous load, and structuring information to support germane processing. This can make tasks more manageable, enhancing competence and autonomy, and supporting intrinsic motivation in the short term by increasing engagement, focus, and task mastery. However, over-reliance on AI may have negative consequences: it can reduce independent problem-solving, critical thinking, and opportunities for meaningful social interaction, undermining autonomy, competence, and relatedness. While AI can enhance motivation by making tasks easier, depending on it too much may reduce long-term intrinsic motivation if it replaces active effort and reduces confidence in completing tasks independently.

Nuvola apps kuser

Take home message: AI is a powerful tool for supporting learning, but its benefits are maximised when it complements, rather than replaces, active cognitive effort, reflective thinking, and collaborative engagement. Long-term motivation and growth depend on maintaining ownership, mastery, and connection in learning.

References

Byers, C. M. (2024). AI-Powered Educational Tools and Their Effect on Student Motivation in Online Learning Environments: A Preliminary Study. Graduate Student Theses, Dissertations, & Professional Papers. 12377. https://scholarworks.umt.edu/etd/12377

Chen, C. H., & Chang, C. L. (2024). Effectiveness of AI-assisted game-based learning on science learning outcomes, intrinsic motivation, cognitive load, and learning behavior. Education and Information Technologies, 29(14), 18621-18642. https://doi.org/10.1007/s10639-024-12553-x

Chen, Y., Wang, Y., Wüstenberg, T. Kizilcec, R., Fan, Y., Li, Y., Lu, B., Yuan, M., Zhang, J., Zhang, Z., Geldsetzer, P., Chen S., & Bärnighausen, T. (2025) Effects of generative artificial intelligence on cognitive effort and task performance: study protocol for a randomized controlled experiment among college students. Trials, 26(1), 244. https://doi.org/10.1186/s13063-025-08950-3

Chiu, T. K. (2024). A classification tool to foster self-regulated learning with generative artificial intelligence by applying self-determination theory: A case of ChatGPT. Educational technology research and development, 72(4), 2401-2416. https://doi.org/10.1007/s11423-024-10366-w

Fan, Y., Tang, L., Le, H., Shen, K., Tan, S., Zhao, Y., Shen, Y., Li, X., & Gasevic, D. (2025). Beware of metacognitive laziness: Effects of generative artificial intelligence on learning motivation, processes, and performance. British Journal of Educational Technology, 56(2), 489-530. https://doi.org/10.1111/bjet.13544

Galindo‐Domínguez, H., Delgado, N., Urruzola, M. V., Etxabe, J. M., & Campo, L. (2025). Using Artificial Intelligence to Promote Adolescents' Learning Motivation. A Longitudinal Intervention From the Self‐Determination Theory. Journal of Computer Assisted Learning, 41(2), e70020. https://doi.org/10.1111/jcal.70020

Gerlich, M. (2025). AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking. Societies, 15(1), 6. https://doi.org/10.3390/soc15010006

Gkintoni, E., Antonopoulou, H., Sortwell, A., & Halkiopoulos, C. (2025). Challenging cognitive load theory: The role of educational neuroscience and artificial intelligence in redefining learning efficacy. Brain Sciences, 15(2), 203. https://doi.org/10.3390/brainsci15020203

Hussain, S. A., Ayub, F., & Ahmed, N. (2025). Cognitive Load Management Through Adaptive AI learning System Implications for Student Focus and Retention. The Critical Review of Social Sciences Studies, 3(3), 701-719. https://doi.org/10.59075/kpfrdv65

Jose B, Cherian J, Verghis A. M., Varghise S. M., & Joseph S. (2025). The cognitive paradox of AI in education: between enhancement and erosion. Frontiers in Psychology, 16, 1550621. https://doi.org/10.3389/fpsyg.2025.1550621

Martin, A. J., Ginns, P., Burns, E. C., Kennett, R., Munro-Smith, V., Collie, R. J., & Pearson, J. (2021). Assessing Instructional Cognitive Load in the Context of Students' Psychological Challenge and Threat Orientations: A Multi-Level Latent Profile Analysis of Students and Classrooms. Frontiers in psychology, 12, 656994. https://doi.org/10.3389/fpsyg.2021.656994

Martin, A. J., Collie, R. J., Kennett, R., Liu, D., Ginns, P., Sudimantara, L. B., Dewi, E. W., & Rüschenpöhler, L. (2025). Integrating generative AI and load reduction instruction to individualize and optimize students' learning. Learning and Individual Differences, 121, 102723. https://doi.org/10.1016/j.lindif.2025.102723

Risko, E. F., & Gilbert, S. J. (2016). Cognitive Offloading. Trends in cognitive sciences, 20(9), 676–688. https://doi.org/10.1016/j.tics.2016.07.002

Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68

Sweller, J. (2010). Element interactivity and intrinsic, extraneous, and germane cognitive load. Educational psychology review, 22(2), 123-138. https://doi.org/10.1007/s10648-010-9128-5

Sweller J., Ayres P., Kalyuga S. (2011). Cognitive Load Theory. Springer. 10.1007/978-1-4419-8126-4

Wu, Y. (2023). Integrating generative AI in education: how ChatGPT brings challenges for future learning and teaching. Journal of Advanced Research in Education, 2(4), 6-10. https://doi.org/10.56397/JARE.2023.07.02

Wu, S., Liu, Y., Ruan, M., Chen, S., & Xie, X. Y. (2025). Human-generative AI collaboration enhances task performance but undermines human’s intrinsic motivation. Scientific Reports, 15(1), 15105. https://doi.org/10.1038/s41598-025-98385-2

External links

Cognitive load theory & AI (AI competence)
Self-determination theory (Centre for self-determination theory)