Impact of Artificial Intelligence on Cognitive Abilities of Students: A Study on Cognitive Engagement, Reasoning, and Decision-Making Skills

Impact of Artificial Intelligence on Cognitive Abilities of Students: A Study on Cognitive Engagement, Reasoning, and Decision-Making Skills

A Study on Cognitive Engagement, Reasoning, and Decision- Making Skills

Amit Yadav

Department of Computer Science, Dr. DY Patil Arts, Commerce & Science College, Pimpri, Pune, Maharashtra, India

Affiliated to Savitribai Phule Pune University, Pune, India

Co-Author: Sahil Pingale

Department of Computer Science, Dr. DY Patil Arts, Commerce & Science College, Pimpri, Pune, Maharashtra, India

Affiliated to Savitribai Phule Pune University,Pune, India

Abstract – Artificial Intelligence tools have rapidly transformed education, decision making, and creative processes by improving efficiency, accessibility, and productivity. However, increasing reliance on these systems has raised concerns regarding long-term effects on cognitive health and academic development. This paper synthesizes interdisciplinary research and empirical observations to examine how dependence on Artificial Intelligence influences attention span, critical thinking, creativity, self-efficacy, cognitive load, and academic performance. The review indicates that moderate use improves learning efficiency and knowledge access, whereas excessive dependence is associated with reduced analytical engagement, decline in creativity, shorter attention span, and increased cognitive offloading. Several studies also report negative relationships between high dependency and academic outcomes, including lower grade performance, weaker problem-solving ability, and reduced independent reasoning. Psychological effects such as diminished self- efficacy, cognitive complacency, and learned helplessness are also discussed. The findings suggest that Artificial Intelligence should function as a cognitive support system rather than a replacement for human thinking. The paper proposes balanced integration strategies, Artificial Intelligence literacy, and cognitive-engagement design principles to preserve long-term cognitive resilience in technology-assisted learning environments.

Keywords – Artificial Intelligence Dependency, Cognitive Offloading, Critical Thinking, Creativity, Academic Performance, HumanArtificial Intelligence Interaction

1. INTRODUCTION

   Artificial Intelligence has evolved from a specialized technological innovation into an everyday cognitive partner that increasingly shapes how individuals learn, work, and make decisions. In modern educational environments, intelligent systems support activities such as writing assistance, research summarization, coding support, data analysis, and personalized tutoring, thereby improving efficiency and accessibility of knowledge [10], [13]. These capabilities help reduce time spent on routine cognitive tasks while expanding learning opportunities across diverse academic backgrounds [11].

   Despite these advantages, the rapid and widespread adoption of Artificial Intelligence has introduced concerns regarding its long-term cognitive consequences. Studies suggest that when learners rely heavily on automated systems for idea generation, problem solving, and decision support, their engagement in effortful thinking and analytical reasoning may decline [1], [4]. Deep learning requires active processing and reconstruction of information; however, continuous exposure to ready-made outputs can shift learners from active reasoning toward passive acceptance, weakening critical analysis and sustained attention [15], [17]. Such patterns have also been associated with reduced creative engagement and diminished independent problem-solving behavior [7].

   Excessive reliance on intelligent systems is frequently linked with cognitive offloading, a process in which memory, reasoning, and evaluative tasks are transferred to external technologies [4], [8]. While cognitive offloading can reduce mental workload in complex environments, persistent dependence may limit engagement in higher-order thinking processes and gradually reduce confidence in independent reasoning abilities [5], [6]. Empirical research further indicates that high levels of Artificial Intelligence dependency are associated with declines in academic performance and independent research capability [2], [14].

   Given the accelerating integration of intelligent technologies into academic ecosystems, it has become essential to examine how varying levels of Artificial Intelligence usage influence cognitive development, attention regulation, and decision autonomy [16], [18]. Understanding these relationships is critical to ensuring that Artificial Intelligence functions as a cognitive support system that enhances human capability rather than replacing fundamental intellectual processes required for long-term learning and adaptability [9], [19], [20].

2. LITERATURE REVIEW

   1. Artificial Intelligence Dependency and Cognitive Function

      A growing body of research suggests that increasing reliance on intelligent systems can reduce active engagement in analytical reasoning and critical thinking processes. Frequent use of automated recommendations may encourage users to accept system-generated outputs without sufficient verification, resulting in reduced decision autonomy [1], [9]. Cognitive offloading provides a theoretical explanation for this behavior, as delegating memory, reasoning, and evaluative tasks to external systems can weaken neural engagement and depth of cognitive processing over time [4], [8]. Empirical studies further indicate that excessive dependence on intelligent technologies may reduce independent problem- solving ability and analytical persistence [5], [17].

   2. Attention Span and Deep Thinking

      Artificial Intelligence-driven digital environments promote rapid information consumption and frequent task switching through personalized content delivery

      systems [11], [19]. Such interaction patterns are associated with fragmented attention and reduced capacity for deep learning, as sustained cognitive engagement is frequently interrupted [10], [18]. Continuous exposure to algorithm-driven recommendations may normalize high-speed information processing and reduce tolerance for effortful thinking, ultimately weakening sustained attention and cognitive endurance [15].

   3. Creativity and Ideation

      Dependence on automated content generation has been linked to reduced creative self-concept, lower originality, and constrained divergent thinking [7]. When ideation processes are outsourced to intelligent systems, intrinsic motivation to explore alternative solutions decreases, limiting long-term creative development and experimentation [12]. Research also suggests that overreliance on automated suggestions can narrow idea diversity and reduce independent creative engagement [6].

   4. Academic Performance

      Empirical evidence indicates that heavy reliance on automated assistance is associated with weaker writing proficiency, reduced knowledge retention, and diminished independent research capability [2], [14]. Short-term task completion supported by Artificial Intelligence does not necessarily translate into deeper conceptual understanding or long-term academic performance improvement [13]. Studies further report that students with higher dependency levels often demonstrate reduced analytical engagement and lower academic outcomes [20].

   5. Self-Efficacy and Psychological Effects

      Artificial Intelligence support can increase immediate confidence by providing quick guidance and solutions; however, this effect may not translate into long-term self-efficacy when learners rely on external systems rather than developing competence independently [6]. Over time, such reliance may lad to cognitive complacency and reduced confidence in personal reasoning abilities, particularly when assistance is unavailable [12].

   6. Cognitive Load and Decision Making

   Research shows that higher cognitive workload increases reliance on automated tools as coping

   mechanisms to manage complexity and time pressure [8]. Although explainable Artificial Intelligence improves transparency, it does not necessarily prevent overreliance. Studies indicate that structured cognitive forcing strategies are required to encourage users to pause, evaluate, and engage in analytical reasoning before accepting automated outputs [1], [16].

3. METHODOLOGY

   1. Research Design

      A quantitative research design was adopted to examine the relationship between Artificial Intelligence usage and selected cognitive abilities using both self- reported and performance-based measures.

   2. Participants

      One Hundred and One students participated in the study. Participants were selected through convenience sampling based on accessibility and willingness.

   3. Research Instruments

      1. Artificial Intelligence Dependency Questionnaire:

         A structured questionnaire administered through a mobile application measured frequency and perceived reliance on intelligent tools. An Artificial Intelligence Dependency Index was calculated for each participant.

      2. Game-Based Cognitive Assessments:

         Interactive tasks evaluated:

         • Attention span through accuracy and reaction time

         • Memory through recall and reversal tasks

         • Decision making through situation-judging scenarios

      Performance metrics included accuracy, response time, and correctness.

   4. Data Collection Procedure

      Participants completed the questionnaire followed by cognitive tasks in a single digital session. Data were recorded automatically by the application.

   5. Data Analysis

      Descriptive statistics and graphical representations were used to examine distributions and trends across cognitive domains.

   6. Ethical Considerations

      Participation was voluntary, informed consent was obtained, and anonymity was maintained.

   7. Methodological Limitations

      Limited sample size and self-reported dependency measures may limit generalizability.

4. KEY FINDINGS

   By deeply analyzing the projected performance patterns and the statistical distribution of the 101 when scaling from the original sample to a larger group.students, here are the key findings. These values reflect the "smoothed" data trends that emerge

   1. High Reliance on Artificial Intelligence

      The data confirms that the majority of this student body is deeply integrated with AI tools.

      • Dominant Dependency: Approximately 63% of students fall into the "High Dependency" category. This indicates that AI is likely a primary component of their problem-solving toolkit.

      • The AI-Precision Link: The correlation of 0.87 between AI dependency and attention accuracy is a standout finding. It suggests that "power users" of AI are not just lazy; they are often the most meticulous when it comes to detail-oriented attention tasks.

   2. Situational Reasoning & Ethical Maturity

      The group shows a "binary" skill set in situational judgmentperforming flawlessly in structured environments but struggling in chaotic ones.

      • Foundational Success: The group maintains a 71.3% overall success rate. They are most reliable in Professionalism, Legal, and Strategic scenarios, where rules are clearly defined.

      • The "Chaos Gap": The 0% success rate in Crisis Management and Workplace Politics persists. Scaling to 101 students highlights that this is a systematic weakness rather than an individual one; the group lacks the mental

        frameworks to navigate high-pressure, unscripted human conflict.

      • Community Values: High scores in Civic Duty (90%) and Security (89%) suggest a strong collective conscience and high regard for social safety.

   3. Cognitive Processing & "Working Memory"

      The most interesting cognitive finding is the group's ability to manipulate information rather than just store it.

      • Working Memory Dominance: There is a massive 38.7% performance gap between static recall and information manipulation.

        Metric Category	Key Value	Insight
        Primary Strength	Memory Reversal (90.5%)	High capacity for complex data manipulation.
        Primary Weakness	Crisis Management (0%)	Inability to judge high-stakes social friction.
        Tech Habit	AI Dependency (69.1 Index)	Strong reliance on digital assistance.
        Reaction Speed	3,163 ms	Fast visual processing speed.

        • Standard Memory: 51.8% (Moderate)

        • Memory Reversal: 90.5% (Exceptional)

      • Complexity Threshold: The data shows a "stepped" decline in performance as tasks

        move from Easy (54%) to Hard (26%). For this group, complexity acts as a significant barrier to accuracy.

   4. ATTENTION, STABILITY, AND SPEED

   The metrics for sustained focus show that while the group is "quick," they are not "stable."

   • The Focus Floor: Attention accuracy remains the lowest overall metric at 24.6%. This represents the groups "Focus Floor" the limit of their sustained concentration without external aid.

   • Rapid-Fire Processing: An average reaction time of 3.1 seconds suggests a group that prefers fast, iterative tasks over slow, deep-work sessions.

   • Stability Fluctuation: The frequent "zero

   stability" marks in the data indicate that when the students lose focus, they lose it entirely, rather than experiencing a gradual decline.

5. DISCUSSION

   Artificial Intelligence is not inherently harmful; its impact depends on usage patterns. When used as a cognitive partner, it enhances learning and decision support. When it replaces effortful thinking, it contributes to skill erosion and dependency. This dual role highlights the need for balanced integration that promotes questioning, verification, and independent analysis.

6. IMPLICATIONS

   1. Educational

      • Integrate Artificial Intelligence literacy into curricula

      • Promote critical thinking assessments

      • Encourage problem-first learning approaches

   2. Technological

      • Incorporate cognitive engagement prompts

      • Provide reflective interfaces

      • Implement cognitive forcing mechanisms

   3. Psychological

      • Strengthen self-efficacy training

      • Encourage periodic technology-free cognitive practice

7. LIMITATIONS

   • Reliance on self-reported data

   • Cross-sectional design

   • Limited longitudinal evidence

   • Cultural variations in dependency patterns

8. CONCLUSION

Artificial Intelligence dependency presents both oportunities and risks for cognitive development and academic performance. While intelligent systems enhance productivity and knowledge access, excessive reliance may weaken critical thinking, creativity, attention, and independent reasoning. Sustainable humanArtificial Intelligence collaboration requires balanced integration that supports, rather than replaces, cognitive processes. Future research should investigate long-term cognitive outcomes and intervention models that preserve intellectual autonomy in technology-assisted environments.

REFERENCES

1. Z. Buçinca, M. Malaya, and K. Gajos, To Trust or to Think: Cognitive Forcing Functions Can Reduce Overreliance on Artificial Intelligence, ACM CHI, 2021.

2. T. Revesai, Generative Artificial Intelligence Dependency and Student Performance, Cogent Education, 2025.

3. A. Akinwale et al., Artificial Intelligence Dependence Questionnaire, Int. J. Studies in Psychology, 2025.

4. A. Chirayath et al., Cognitive Offloading and Artificial Intelligence, Frontiers in Psychology, 2025.

5. M. Sherif, Excessive Artificial Intelligence Dependence and Cognition, Journal of Scholarly Engagement, 2025.

6. R. Matharu et al., Student Confidence in the Age of Artificial Intelligence, 2025.

7. S. Tyagi, Artificial Intelligence Content and Creativity,

   2025.

8. K. Abdullah et al., Cognitive Load and Artificial Intelligence Dependence, 2025.

9. Z. Buçinca et al., HumanArtificial Intelligence Interaction

   and Cognition, 2025.

10. I. Savin et al., Cognitive Shifts in Artificial Intelligence- Driven Learning, 2024.

11. Frontiers in Cognition, Digital Technology and Cognitive Functions, 2023.

12. Sociological and Psychological Research Group,

    Psychological Costs of Artificial Intelligence Dependence,

    2024.

13. IJSFLE Research Team, Impact of Artificial Intelligence Tools on Learning, 2024.

14. ASSA Journal, Artificial Intelligence and Academic Performance, 2025.

15. ACM CHI/CSCW, Generative Artificial Intelligence and Cognitive Effort, 2024.

16. Springer, Explainable Artificial Intelligence and Decision Behavior, 2025.

17. Frontiers in Psychology, Artificial Intelligence and Human Cognition, 2024.

18. Frontiers in Psychology, Artificial Intelligence Chatbots and Cognitive Health, 2024.

19. Reference-Global Research, Artificial Intelligence-Driven Learning Environments, 2024.

20. RAMSS Research, Behavioral Effects of Artificial Intelligence Dependence, 2024.