See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/377530398

Unraveling the Complexity: A DEMATEL Analysis of the Negative Impact of

Artificial Intelligence (AI) Adoption among Students in Higher Education

Article in Journal of Intelligent Systems and Internet of Things · January 2024

DOI: 10.54216/JISIoT.110203

CITATION READS

1 3,177

3 authors:

Zahari Md Rodzi Faisal Al-Sharqi

Universiti Teknologi MARA University of Anbar
49 PUBLICATIONS 144 CITATIONS 76 PUBLICATIONS 577 CITATIONS

SEE PROFILE SEE PROFILE

Ashraf Al-Quran
King Faisal University
72 PUBLICATIONS 786 CITATIONS

SEE PROFILE

All content following this page was uploaded by Faisal Al-Sharqi on 19 January 2024.

The user has requested enhancement of the downloaded file.

 Journal of Intelligent Systems and Internet of Things
Vol. 11, No. 02, PP. 30-41, 2024
DOI: https://doi.org/10.54216/JISIoT.110203

Unraveling the Complexity: A DEMATEL Analysis of the

Negative Impact of Artificial Intelligence (AI) Adoption
among Students in Higher Education

Zahari Md Rodzi 1,2, *, Wan Normila Mohamad3, Zhang Lu4, Faisal Al-Sharqi5, Rawan A. shlaka6,
Ashraf Al-Quran7, Ali M. Alorsan Bany Awad8
1
Pusat Pengajian Matematik, Universiti Teknologi Mara Cawangan Negeri Sembilan, Kampus Seremban,
70300 Seremban, Malaysia.
2
Fakulti Sains dan Teknologi, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor Darul Ehsan.
3
Faculty of Business Management Universiti Teknologi MARA Cawangan Negeri Sembilan, Kampus
Seremban.
4
College of Communication, Zaozhuang University, Zaozhuang, 277000, Shandong, China,
5
Department of Mathematics, Faculty of Education for Pure Sciences, University Of Anbar, Ramadi,
Anbar, Iraq.
6
College of Pharmacy, National University of Science and Technology, Dhi Qar, Iraq
7
Basic Sciences Department, Preparatory Year Deanship, King Faisal University, Al-Ahsa 31982, Saudi
Arabia
8
Deanship of Development and Quality Assurance, King Faisal University, Al-Ahsa 31982, Saudi Arabia
Emails: zahari@uitm.edu.my; surinazri0084@gnail.com; zzah.abdrahman.125@gmail.com;
zlu3276@gmail.com; faisal.ghazi@uoanbar.edu.iq; rawan-a.shlaka@nust.edu.iq; aalquran@kfu.edu.sa;
abanyawad@kfu.edu.sa
*Corresponding Author: zahari@uitm.edu.myAbstract

Abstract

This research employs DEMATEL analysis as a methodological approach to thoroughly examine the adverse
consequences of implementing Artificial Intelligence (AI) among students enrolled at Universiti Teknologi
MARA (UiTM) Negeri Sembilan, Malaysia. The analysis encompasses three distinct professional cohorts:
student representatives, academic staff, and upper management. Through a systematic analysis of causal
relationships between multiple factors, this study aims to identify and prioritize the fundamental elements
contributing to the negative consequences associated with integrating artificial intelligence. The prominence
of privacy and security concerns as a causal factor highlights the importance of implementing strong data
protection measures and adhering to ethical practices related to AI. Furthermore, various factors connected
with personal disconnection, restricted adaptability, dependance on technology, and insufficient emotional
intelligence influence the adverse outcomes of artificial intelligence implementation among students. The
results underscore the necessity of implementing focused interventions and strategies to tackle these difficulties
and guarantee a harmonious and advantageous integration of artificial intelligence in students' educational
journeys. Higher education institutions can effectively harness the advantages of AI while ensuring their
students' welfare and educational achievements by recognizing and proactively addressing any potential
limitations.

Received: August 01, 2023 Revised: November 14, 2023 Accepted: January 04, 2024

Keywords: DEMATEL analysis; Artificial Intelligence; Higher Education; Student Welfare; Data
Protection.

30
 1. Introduction

AI has rapidly gained prominence in higher training, presenting revolutionary answers that intention to
convert coaching and gaining knowledge of procedures [1]. But the adoption of AI in academic settings has
added forth concerns regarding its ability negative effect on university students' instructional studies and
effects. it's far critical to comprehensively study the elements contributing to this negative effect and
recognize their interdependencies to increase powerful techniques for mitigation.
Preceding research has shed mild on numerous poor aspects of AI adoption in education. privateness and
security concerns have been diagnosed as key factors that preclude a hit implementation of AI in
instructional settings [2]. Restricted customization of AI era, leading to a standardized academic experience,
has additionally been highlighted as a capacity difficulty [3]. Emotional disconnection bobbing up from
elevated reliance on AI-based totally structures, which might also reduce personal interplay and human
connection, has been diagnosed as a concerning outcome [4]. Moreover, era dependence and the capability
for improved display screen time because of AI adoption pose risks to critical thinking skills and usual
well-being [5].
To cope with the want for a complete analysis of the negative impact of AI adoption in higher education,
this has a look at employs the DEMATEL approach. The DEMATEL method permits a scientific
assessment of causal relationships amongst various factors contributing to the poor effect of AI adoption.
By applying this method, the study aims to perceive the distinguished factors and their interdependencies,
offering a holistic expertise of the complicated motive-and-impact relationships inside the context of AI
adoption in better training.
The goal of this study is to analyze the poor effect of AI adoption in higher education the use of the
DEMATEL method. with the aid of analyzing the relationships amongst factors together with privacy and
protection concerns, limited customization, emotional disconnection, era dependence, and extended screen
time, the study aims to find the underlying mechanisms and dynamics contributing to AI adoption's poor
effects in education. Ultimately, the findings will tell the improvement of centered interventions and
strategies to mitigate the terrible impact, ensuring a balanced and useful integration of AI in better education
settings.

2. Literature Review

The issue of personal disconnection springing up from the improved integration of AI in education has
garnered attention within the literature. As AI-primarily based structures become more every day, there
may be a subject that students may additionally revel in a reduced sense of personal interplay and human
connection, which can be vital for powerful gaining knowledge of studies [6]. The reliance on AI era may
additionally create a barrier to fostering significant pupil-instructor and peer-to-peer interactions, impacting
social and emotional improvement [7]. This lack of personal connection can lead to decreased engagement,
motivation, and delight inside the instructional technique [8]. Addressing this task requires finding a
stability between AI integration and maintaining the critical human element in education.
The constrained customization of AI generation in academic contexts has been identified as a extensive
concern. while AI offers the potential for personalized learning reviews, the rigid nature of a few AI systems
may additionally prevent character student wishes and preferences [9]. students may have various studying
patterns, abilties, and hobbies that require tailored methods. however, the dearth of customization options
can result in a standardized academic revel in that fails to address the specific requirements of each learner
[10]. To absolutely leverage the blessings of AI, it's miles crucial to broaden adaptable and bendy AI
structures that cater to person mastering wishes and sell personalised academic pathways. the supply of
erroneous or deceptive facts by way of AI systems is another regarding factor. whilst AI technology can
enhance statistics retrieval and evaluation, there may be a danger of erroneous or biased outputs [11]. This
underscores the importance of ensuring the accuracy, reliability, and transparency of AI algorithms and
statistics assets [12]. erroneous information can cause inaccurate mastering outcomes, misinformation
propagation, and erosion of believe in AI-based educational tools [13]. To mitigate this, ongoing tracking,
rigorous validation techniques, and continuous development of AI structures' content and hints are crucial.
Privateness and security worries surrounding AI adoption in university had been significantly discussed.
AI structures frequently rely on collecting and studying great quantities of touchy pupil information, raising
questions about facts protection, consent, and ability breaches [14]. Safeguarding student privacy is critical
to keep believe in AI technology and make certain compliance with relevant rules. instructional institutions

31
and policymakers want to set up robust privateness regulations, comfy information control practices, and
ethical recommendations for AI implementation in university [15]. The dependence on AI technology can
affect university students' critical wondering capabilities and trouble-solving capabilities. Overreliance on
AI structures may also diminish university students' independent questioning and creativity, as AI
frequently provides geared up-made answers [16]. students should be recommended to broaden their
cognitive and analytical capabilities, even within the presence of AI gear, to foster deeper know-how and
domesticate vital talents for the future [17].
AI systems' limited emotional intelligence poses challenges to the development of empathy and
interpersonal abilities amongst students. Emotional intelligence, which includes recognizing and
responding to emotions, is an important issue of holistic education [18]. AI's inability to completely
recognise and respond to human emotions can prevent the cultivation of emotional intelligence amongst
university students, which is vital for powerful verbal exchange, collaboration, and nicely-being [19]. the
mixing of AI era in training may also make contributions to extended screen time for both students and
educators. immoderate display time has been associated with numerous bodily and mental health dangers,
including sedentary conduct, eye stress, sleep disturbances, and decreased social interaction [20]. it is
critical to strike a balance between utilising AI tools and promoting wholesome display screen behavior by
way of incorporating offline activities, bodily movement, and screen-time management techniques [21-
26,37].
DEMATEL method changed into at the beginning advanced via the technological know-how and Human
Affairs Programme (SHAP) of the Battelle Memorial Institute of Geneva between 1972 and 1976 [22]. The
research employed a methodology primarily based on graph idea to advantage talent and illustrate the
complex relationships among different variables the usage of a visible illustration [23]. The usage of the
DEMATEL method may facilitate the process of prioritizing factors thru the analysis of the visual
interconnections among entities and their respective groups. This analysis helps the discernment of the
nature of the connection among variables and the importance of their have an impact on other variables.
The DEMATEL technique has won significant traction in diverse fields together with the control of know-
how, advertising processes, structures for manage, and safety issues in recent times.
3. METHODOLOGY
The present study's methodology chapter delineates the systematic approach employed to examine the
adverse effects associated with the integration of AI in higher education settings, specifically among
students. The study is carried out in four separate phases to guarantee a thorough and rigorous examination.
Phase 1: Literature Review and Identification of Negative Impact Criteria
The have a look at initiates by using accomplishing a complete literature evaluation to examine the present-
day corpus of information approximately the negative effect of AI implementation amongst university
students in better education. The literature overview has recognized several criteria that are associated with
the detrimental effects of incorporating AI technology in instructional environments. these standards
encompass an extensive spectrum of formerly mentioned concerns as viable limitations and boundaries.
The complete examination of existing literature paperwork the foundation for determining the standards so
one can be assessed in next tiers. The standards that have been diagnosed are complete in nature and embody
the various dimensions of the negative results which can arise from the adoption of synthetic intelligence
amongst university students. thru a complete examination and integration of previous scholarly
investigations, this section ensures that the modern studies are constructed upon a robust basis of mounted
knowledge and enlightens the following ranges of the take a look at. the realization of this stage is desk 1,
which presents an in-depth stock of the destructive outcomes associated with the combination of AI among
university students in institutions of higher mastering. The table provided herein offers a complete summary
of the numerous standards to be able to be difficulty to further evaluation within the subsequent ranges of
the research.
Phase 2: Expert Assessment and Criteria Selection
Student, academic staff, and higher-level administrative representatives are chosen to evaluated criteria in
section 2. The direct score method are used to assign ratings to each poor effect criteria on a scale ranging
from 1 to 10. On this scale, a grade of 1 represents "Very not crucial" while a rating of 10 represents "Very
vital". a fixed threshold value of 7.5 or above become decided on at some stage in the score manner. Any
poor impact criteria that exceeded this degree was considered very tremendous and selected for this studies.
This method enabled a targeted of the main factors which have a tremendous effect on the adoption and use
of AI among university students.

32
Phase 3: Pairwise Comparison of Selected Criteria
Phase 3 comprises paired comparisons on a linguistic scale from 1 to 5 to determine the degree of
connection between the criteria. The scale includes a range of categories, including "Unrelated" (1),
"Slightly Unrelated" (2), "Moderately Unrelated" (3), "Related" (4), and "Very Related" (5). This practise
includes experts' subjective assessments of criterion relationships. The researchers use linguistic
terminology to gain a thorough understanding of how different factors are connected. This helps them
identify the most important factors that contribute to the negative effects of implementing AI among
students in higher education. The Phase 4 DEMATEL study uses the data to provide treatments and
strategies to overcome artificial intelligence adoption obstacles.
Phase 4: DEMATEL Analysis Execution
The DEMATEL technique is a systematic strategy used to investigate causal links and evaluate the
importance of each criterion. The DEMATEL analytical process involves many steps:

Step 1: In the first stage, a direct connection matrix is formed, represented as xijk , where each element
represents an integer score k supplied by an expert. The metric, as specified in Equation (1), offers a
numerical assessment of the impact of criteria i on criterion j inside a matrix of dimensions n×n. The phrase
may be rewritten as:

1 H k
a jj =  xij
L k =1
(1)

L: The number of experts

Step 2: The second step involves normalizing the matrices representing direct influence. The normalized
direct-relation matrix R is derived from matrix C using equations (2) and (3).

S =kA (2)

 
1 1
k = min  ,  i, j  (1, 2,..., n) (3)
 max 
1 i  n  j =1  ij  1 j  n  i =1  ij  
   
n n
a max a


Step 3: The next step in the process involves the development of a comprehensive relations matrix. The
computation of the total-relation matrix W involves the utilization of formula (4), wherein the Identity
Matrix I is employed after the normalization of the direct-relation matrix R.

M = S (I − S)
−1
(4)

Step 4: Implementing the concepts of C+R and R–C is exemplified in equations (5) to (7) for matrix M,
where R denotes the summation of columns and C denotes the summation of rows. Criteria that demonstrate
positive R-C values have a greater impact on the other criteria. The term "dispatchers" is frequently
employed to designate these criteria. Criteria with lower levels of R–C are more vulnerable to external
influences. The entities are commonly referred to as "receivers." In contrast, the summation of variables R
and C represents the degree of correlation between a single criterion and the remaining variables.

M =  mij  i, j  (1, 2,..., n) (5)

n n

Ri =  j =1  mij  =  mi n1
n
n1
(6)

Ci =  i =1  mij  =  m j 
n
1 n 1 n
(7)

33
Step 5: Obtain the "matrix of internal dependencies" and the "map of effect relationships." The dataset was
mapped using the values of (R+C, R–C). The parameter commonly referred to as the threshold value is
utilized to indicate the level of interdependence that exists within a given set of criteria.
4. Results And Discussion
This chapter summarizes the studies on AI integration's negative impacts on higher education students. The
chapter begins with Table 1, a thorough list of negative effect criteria with brief explanations and average
ratings.
Table 1: List all negative impact.

Negative Brief Description Rating Selected

Impact Value Criteria
Personal Reduced sense of personal interaction and human Selected
Disconnection connection, impacting social and emotional 8.54 (N1)
development.
Limited Lack of adaptability in AI systems, hindering Selected
Customization personalized learning experiences tailored to 7.98 (N2)
individual needs.
Lack of The absence or limited presence of emotional Selected
Emotional intelligence in AI systems hampers the development 7.56 (N3)
Intelligence of empathy and interpersonal skills among students.
Privacy and Concerns about data protection, consent, and Selected
Security potential breaches in student data collection and use. 8.56 (N4)

Technology Excessive reliance on AI may hinder students' critical Selected

Dependence thinking and problem-solving abilities. 8.13 (N5)

Wrong The provision of inaccurate or biased information by Selected

Information AI systems leads to misguided learning outcomes. 8.01 (N6)

Increase Screen The integration of AI technology may contribute to Not

Time increased screen time for students and educators. 7.89 Selected
(N7)
Standardization Lack of customization in AI systems may result in a Not
of Education standardized educational experience, ignoring diverse 7.48 Selected
needs.
Ethical Encountering ethical dilemmas related to AI use, Not
Dilemmas such as plagiarism concerns or AI-generated content 7.47 Selected
ownership.
Loss of Human Excessive use of AI may reduce opportunities for Not
Interaction meaningful human interactions and mentorship. 7.44 Selected

Skill Gaps Overemphasis on AI-integrated learning may neglect Not

the development of essential skills not addressed by 7.38 Selected
AI tools.
Dependency on Overreliance on AI-generated recommendations may Not
AI for Decision- hinder students' ability to make informed decisions. 6.98 Selected
Making
Disengagement Overusing AI tools may lead to disengagement and Not
disinterest in the learning process among some 6.97 Selected
students.
Technological Continuous exposure to AI and technology may Not
Overwhelm cause students to feel overwhelmed and stressed. 6.54 Selected

Data Misuse Mishandling or misuse of student data by AI systems Not

and Breach can lead to privacy breaches and trust issues. 6.52 Selected

Figure 1 depicts the chosen parameters pertaining to the adverse effects of AIon students.

34
 Figure 1: The selected criteria on the negative impact of AI among students

Tables 2, 3, and 4 are fundamental components of the DEMATEL analysis, for assessing the
interrelationships and significance of decision criteria. Table 2, referred to as the Average Initial Direct
Relationship Matrix, A, captures the expert-assessed influence levels between criteria N1 to N7. Table 3,
known as the Normalized Direct Relation Matrix, D, standardizes these relationships to a common scale
for relative comparisons. Lastly, Table 4, the Total Relationship Matrix T, synthesizes direct and indirect
influences among criteria, providing a comprehensive perspective on their importance in the decision-
making process.

Table 2: Average initial direct relationship matrix A

N1 N2 N3 N4 N5 N6 N7
N1 1.0000 3.0000 2.6667 2.3333 4.3333 2.0000 3.0000
N2 4.0000 1.0000 2.3333 2.6667 2.6667 2.0000 4.3333
N3 2.6667 3.6667 1.0000 2.3333 2.0000 4.0000 2.0000
N4 4.6667 4.6667 4.3333 1.0000 5.0000 4.0000 4.0000
N5 3.3333 3.0000 2.3333 3.3333 1.0000 4.0000 1.6667
N6 4.3333 3.3333 2.3333 4.0000 2.0000 1.0000 3.6667
N7 2.0000 2.3333 4.0000 2.3333 2.3333 4.0000 1.0000

Table 3: Normalized direct relation matrix D.

N1 N2 N3 N4 N5 N6 N7
N1 0.0361 0.1084 0.0964 0.0843 0.1566 0.0723 0.1084
N2 0.1446 0.0361 0.0843 0.0964 0.0964 0.0723 0.1566
N3 0.0964 0.1325 0.0361 0.0843 0.0723 0.1446 0.0723
N4 0.1687 0.1687 0.1566 0.0361 0.1807 0.1446 0.1446
N5 0.1205 0.1084 0.0843 0.1205 0.0361 0.1446 0.0602
N6 0.1566 0.1205 0.0843 0.1446 0.0723 0.0361 0.1325
N7 0.0723 0.0843 0.1446 0.0843 0.0843 0.1446 0.0361

Table 4: Total relationship matrix T

N1 N2 N3 N4 N5 N6 N7
N1 0.3052 0.3599 0.3261 0.3064 0.3849 0.3293 0.3451
N2 0.4116 0.3006 0.3266 0.3228 0.3421 0.3369 0.3983
N3 0.3590 0.3761 0.2624 0.3021 0.3033 0.3823 0.3130
N4 0.5527 0.5337 0.4880 0.3664 0.5175 0.5112 0.4931

35
 N5 0.3993 0.3733 0.3247 0.3494 0.2874 0.3998 0.3178
N6 0.4545 0.4085 0.3522 0.3897 0.3493 0.3269 0.4067
N7 0.3392 0.3379 0.3663 0.3051 0.3146 0.3903 0.2771

Table 5 presents the outcomes of the DEMATEL analysis, which depict the comparative significance of
the chosen criteria and their respective contributions to the overall adverse influence.
Table 5: Prominence and Net Cause/Effects for Negative Impact of AI

Criteria R C R+C R-C Identify Ranking

N1- Personal Disconnection 2.3569 2.8217 5.1785 -0.4648 Effect 3
N2- Limited Customization 2.4389 2.6901 5.1290 -0.2512 Effect 4
N3- Lack of Emotional Intelligence 2.2983 2.4462 4.7445 -0.1478 Effect 7
N4- Privacy and Security 3.4627 2.3420 5.8047 1.1206 Cause 1
N5- Technology Dependence 2.4517 2.4991 4.9508 -0.0474 Effect 5
N6- Wrong Information 2.6877 2.6766 5.3643 0.0111 Cause 2
N7- Increase Screen Time 2.3306 2.5511 4.8817 -0.2205 Effect 6

Table 5, offer insights into the relative significance of every criterion and its contribution to the overall
terrible effect. Among the recognized standards, N4 demonstrates the very best prominence and internet
purpose/effect with a rating of five.8047, indicating it as a big cause thing. The difficulty for data safety
and privateness acts as a barrier to the successful implementation of AI in educational settings. This locating
emphasizes the significance of addressing privateness and safety troubles to make certain the accountable
and ethical use of AI in higher training. The standards with super negative results, even though to a quite
lesser quantity, encompass N1 with a score of five.1785, N2 with a score of 5.1290, and N5 with a score of
4.9508. These factors spotlight the capability demanding situations of reduced personal interplay,
standardized instructional reports, and decreased crucial wondering abilities because of AI adoption.
On the opposite hand, N6 reveals a mild prominence and internet reason/impact with a score of 5.3643,
indicating it’s had an effect on as a reason thing. The absence or restrained presence of emotional
intelligence in AI systems hinders the improvement of empathy and interpersonal capabilities among
university students, negatively impacting their ordinary academic revel in. Factors which include N3 with
a score of 4.7445 and N7 with a rating of 4.8817 show a extraordinarily decrease net purpose/impact. While
those factors still make contributions to the negative effect of AI adoption, there influence is rather less
prominent compared to the opposite standards.
These findings from the DEMATEL analysis highlight the significance of addressing privateness and
protection issues in AI adoption at the same time as additionally emphasizing the importance of mitigating
factors related to non-public disconnection, confined customization, era dependence, loss of emotional
intelligence, incorrect records, and increased screen time. By recognizing these key elements and their
causal relationships, stakeholders can expand interventions and techniques to triumph over the negative
impact of AI adoption in higher training efficiently. The new overall relation matrix for the chosen standards
is provided in Table 6, presenting extra insights into the relationships and interdependencies several of the
elements.
Table 6: Total relation matrix for Negative Impact of AI
N1 N2 N3 N4 N5 N6 N7
Individual N1 0 0 0 0 1 0 0 criterion
network maps, N2 1 0 0 0 0 0 1 featured
in Figure 2, focus N3 0 1 0 0 0 1 0 on the
relationships of a N4 1 1 1 0 1 1 1 particular
criterion within the
N5 1 1 0 0 0 1 0
context of AI's dire
impact on N6 1 1 0 1 0 0 1
education. These N7 0 0 0 0 0 1 0 detailed

36
visualizations primarily rely on an attached threshold fee of 0.3679 to depict how each criterion interacts
with others. In comparison, Figure 3 presents a comprehensive view by displaying a general community
map showing all combined connections between criteria. Each criterion's importance and the intricate
network of relationships characterising artificial intelligence's harmful consequences in education become
clarified through these visible representations.

Figure 2: Network relationship map for each criterion

Figure 3: Network relationship map.

The whole correlation matrix suggests that privateness and security (N4) is essential, having sturdy
connections with other components. consequently, it is a chief reason of artificial AI's negative results on

37
university. The matrix illustrates that N4 has a sizeable influence on many different elements, including
N1, N2, N3, N5 and N6. The declaration implies that privacy and protection worry influence the use of AI
technologies in education, which affect personal disconnection, adaptability, emotional intelligence,
generation dependence, and data accuracy. The N6 criteria is likewise vital and interconnected. The entire
correlation matrix suggests that N6 impacts N1, N2, N4 and N7. This shows that faulty or prejudiced facts
in AI structures can lessen interpersonal engagement, preclude tailor-made educational reports, enhance
privacy and protection issues, and growth display time for university students and educators.
Han, Buchanan and Mckay [24] added attention to the effect of privacy and security issues associated with
AI technology on university students' sense of personal disconnection, as well as the potential detrimental
consequences at the adaptability and customization of their learning reports. Moreover, the importance of
rectifying inaccurate data in AI systems to assure unique and dependable academic effects changed into
underscored by Scott [25]. N1 exerts a causal effect on N5, whereas N7 exerts a causal have an impact on
N6. This shows that through addressing issues related to private disconnection and the increasing quantity
of time spent on monitors, there may be the capability to mitigate generation dependence and decorate the
accuracy of facts in artificial intelligence systems.

5. MANAGERIAL IMPLICATIONS
The adoption of AI in university provides diverse organizational and managerial challenges, such as method
alignment, human aid control, and operational practices [1]. To cope with those challenges, educational
institutions should prioritize the implementation of sturdy records protection measures and ethical practices
related to AI [27]. This includes the improvement of stringent privacy rules, comfortable statistics
management practices, and clean suggestions for AI implementation to construct and preserve trust among
university students and different stakeholders. moreover, the findings emphasize the significance of
considering students' views on privacy in generation-stronger guides [28]. Administrator should cope with
university students' privateness worries by way of integrating privateness manipulate measures and
fostering believe in the higher education institution to make certain the ethical use of AI technology [29].
Moreover, the integration of emotional intelligence components into AI systems is important to cope with
the impact of the shortage of emotional intelligence on students' interpersonal competencies and empathy
[30]. Educational institutions must explore ways to integrate emotional intelligence components into AI
structures to decorate the socio-emotional aspects of learning. moreover, selling safety recognition and
addressing privacy worries inside the educational community is critical for responsible AI usage [31]. This
involves offering training and recognition programs to students, school, and staff on responsible AI
utilization, facts safety, and ethical issues to create a more informed and responsible AI consumer
community in the institution.
AI usage in improved education raises ethical and privacy concerns that must be addressed. To ensure AI
information correctness and trustworthiness, educational institutions must prioritize constant monitoring,
rigorous validation methods, and transparency in AI algorithms and facts assets. Continuous evaluation can
provide valuable insights for adapting AI strategies and interventions to effectively tackle increasingly
complex scenarios. Higher education institutions can optimize and enhance their operations by consistently
monitoring the impacts of AI to guarantee that it remains a valuable instrument for instructional purposes.
6. CONCLUSION
In conclusion, this study examined the negative effects of AI adoption in higher education using
DEMATEL analysis. The study revealed numerous key findings that illuminate the complexities of
integrating AI in education, with management consequences. The emergence of privacy and security
concerns as a central issue highlighted the critical nature of solid data protection measures and ethical
conduct in relation to artificial intelligence. To achieve and retain student and stakeholder trust, educational
institutions must prioritize strict privacy regulations, secure data management, and clear AI deployment
guidelines. It is vital to consider the viewpoints of students regarding privacy, as doing so cultivates a
perception of agency and confidence in AI-integrated courses. For the ethical application of AI
technologies, institutions must implement privacy control measures and cultivate trust. Moreover, the
research emphasized the importance of incorporating elements of emotional intelligence into artificial
intelligence systems. Interpersonal competence and empathy, which are essential components of a well-
rounded education, may be impeded by AI's deficiency in emotional intelligence. Therefore, it is
recommended that academic establishments investigate potential methods of integrating emotional
intelligence elements into artificial intelligence systems, thereby augmenting the socio-emotional

38
dimensions of the learning process. The study also stressed the relevance of AI system data accuracy and
reliability. It is critical to implement stringent validation procedures, continuous monitoring, and
transparency measures in AI algorithms and data sources. By adopting this proactive strategy, organizations
can uphold the reliability of AI tools, thereby encouraging conscientious utilization and well-informed
judgment. Together, these results offer significant contributions to the understanding of the intricate terrain
surrounding the implementation of AI in higher education. Educational institutions can effectively navigate
the challenges offered by AI while capitalizing on its benefits, through a steadfast commitment to privacy,
emotional intelligence, and data accuracy. These administrative implications help institutions integrate AI
into students' educational journeys to improve outcomes and experience.

Acknowledgments: This work was supported by the Deanship of Scientific Research, Vice Presidency for
Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia [Grant No. GRANT5443].
Funding: This work was supported by the Deanship of Scientific Research, Vice Presidency for Graduate
Studies and Scientific Research, King Faisal University, Saudi Arabia [Grant No. GRANT5443].

References

[1] N. D. Nguyen, “Exploring the role of AI in education,” London J. Soc. Sci., no. 6, pp. 84–95,
2023, doi: 10.31039/ljss.2023.6.108.
[2] Q. Bu, “Ethical Risks in Integrating Artificial Intelligence into Education and Potential
Countermeasures,” Sci. Insights, vol. 41, no. 1, pp. 561–566, 2022, doi: 10.15354/si.22.re067.
[3] J. Mason, B. E. Peoples, and J. Lee, “Questioning the scope of AI standardization in learning,
education, and training,” J. ICT Stand., vol. 8, no. 2, pp. 107–122, 2020, doi: 10.13052/jicts2245-
800X.822.
[4] L. Gkinko and A. Elbanna, “Hope, tolerance and empathy: employees’ emotions when using an
AI-enabled chatbot in a digitalised workplace,” Inf. Technol. People, vol. 35, no. 6, pp. 1714–1743,
2022, doi: 10.1108/ITP-04-2021-0328.
[5] I. Akulwar-Tajane, K. K. Parmar, P. H. Naik, and A. V. Shah, “Rethinking Screen Time during
COVID-19: Impact on Psychological Well-Being in Physiotherapy Students,” Int. J. Clin. Exp. Med.
Res., vol. 4, no. 4, pp. 201–216, 2020, doi: 10.26855/ijcemr.2020.10.014.
[6] K. Seo, J. Tang, I. Roll, S. Fels, and D. Yoon, “The impact of artificial intelligence on learner–
instructor interaction in online learning,” Int. J. Educ. Technol. High. Educ., vol. 18, no. 1, 2021, doi:
10.1186/s41239-021-00292-9.
[7] A. Guilherme, “AI and education: the importance of teacher and student relations,” AI Soc., vol.
34, no. 1, pp. 47–54, 2019, doi: 10.1007/s00146-017-0693-8.
[8] J. A. Durlak and R. P. Weissberg, “Promoting social and emotional development is an essential
part of students’ education,” Hum. Dev., vol. 54, no. 1, pp. 1–3, 2011, doi: 10.1159/000324337.
[9] R. Tiwari, “The integration of AI and machine learning in education and its potential to personalize
and improve student learning experiences,” Interantional J. Sci. Res. Eng. Manag., vol. 07, no. 02, pp.
1–11, 2023, doi: 10.55041/ijsrem17645.
[10] M. Dimitrova, C. Sadler, A. Murphy, and S. Hatzipanagos, “WestminsterResearch learning
environments . in e-Learning Environments,” Expert Syst., 2003.
[11] B. D. Horne, D. Nevo, J. O’Donovan, J. H. Cho, and S. Adalı, “Rating reliability and bias in news
articles: Does AI assistance help everyone?,” Proc. 13th Int. Conf. Web Soc. Media, ICWSM 2019, no.
Icwsm, pp. 247–256, 2019, doi: 10.1609/icwsm.v13i01.3226.
[12] M. Pushkarna, A. Zaldivar, and O. Kjartansson, Data Cards: Purposeful and Transparent Dataset
Documentation for Responsible AI. 2022.
[13] T. Nazaretsky, M. Cukurova, M. Ariely, and G. Alexandron, “Confirmation bias and trust: Human
factors that influence teachers’ attitudes towards AI-based educational technology,” CEUR Workshop
Proc., vol. 3042, 2021.
[14] “Latham , Annabel ORCID logoORCID : https://orcid.org/0000-0002-8410- 7950 and Goltz , Sean
( 2019 ) A Survey of the General Public ’ s Views on the Ethics of using AI in Education . In : Lecture
Notes in Artificial Intelligence 11625 - AIED 2019 . Lecture,” pp. 0–13, 2019.

39
[15] W. Holmes et al., “Ethics of AI in Education : Towards a Community-Wide Framework,” Int. J.
Artif. Intell. Educ., vol. 32, pp. 504–526, 2022, doi: https://doi.org/10.1007/s40593-021-00239-1.
[16] K. Khare, B. Stewart, and A. Khare, “Artificial Intelligence and the Student Experience: An
Institutional Perspective,” IAFOR J. Educ., vol. 6, no. 3, pp. 63–78, 2018, doi: 10.22492/ije.6.3.04.
[17] O. Kouzov, “The New Paradigms In Education and Support of Critical Thinking with Artificial
Intelligence (AI) Tools,” Serdica J. Comput., vol. 13, no. 1–2, pp. 27–40, 2019, doi:
10.55630/sjc.2019.13.27-40.
[18] M. Estrada, D. Monferrer, A. Rodríguez, and M. Á. Moliner, “Does emotional intelligence
influence academic performance? The role of compassion and engagement in education for sustainable
development,” Sustain., vol. 13, no. 4, pp. 1–18, 2021, doi: 10.3390/su13041721.
[19] O. Oritsegbemi, “Human Intelligence versus AI: Implications for Emotional Aspects of Human
Communication,” J. Adv. Res. Soc. Sci., vol. 6, no. 2, pp. 76–85, 2023, doi: 10.33422/jarss.v6i2.1005.
[20] V. S. Nakshine, P. Thute, M. N. Khatib, and B. Sarkar, “Increased Screen Time as a Cause of
Declining Physical, Psychological Health, and Sleep Patterns: A Literary Review,” Cureus, vol. 14, no.
10, pp. 1–9, 2022, doi: 10.7759/cureus.30051.
[21] C. P. Society, D. Health, and T. Force, “Digital media: Promoting healthy screen use in school-
aged children and adolescents,” Paediatr. Child Heal., vol. 24, no. 6, pp. 402–408, 2019, doi:
10.1093/pch/pxz095.
[22] Jamiatun Nadwa Ismail et al. The Integrated Novel Framework: Linguistic Variables in
Pythagorean Neutrosophic Set with DEMATEL for Enhanced Decision Support. Int. J. Neutrosophic
Sci., vol. 21, no. 2, pp. 129-141, 2023.
[23] Z. bin M. Rodzi et al., “Integrated Single-Valued Neutrosophic Normalized Weighted Bonferroni
Mean (SVNNWBM)-DEMATEL for Analyzing the Key Barriers to Halal Certification Adoption in
Malaysia,” Int. J. Neutrosophic Sci., vol. 21, no. 3, pp. 106–114, 2023.
[24] Ashraf Al-Quran, Faisal Al-Sharqi, Zahari Md. Rodzi, Mona Aladil, Rawan A. shlaka, Mamika
Ujianita Romdhini, Mohammad K. Tahat, Obadah Said Solaiman. (2023). The Algebraic Structures of
Q-Complex Neutrosophic Soft Sets Associated with Groups and Subgroups. International Journal of
Neutrosophic Science, 22 ( 1 ), 60-76.
[25] F. Al-Sharqi, A. Al-Quran, M. U. Romdhini, Decision-making techniques based on similarity
measures of possibility interval fuzzy soft environment, Iraqi Journal for Computer Science and
Mathematics, vol. 4, pp.18–29, 2023.
[26] F. Al-Sharqi, Y. Al-Qudah and N. Alotaibi, Decision-making techniques based on similarity
measures of possibility neutrosophic soft expert sets, Neutrosophic Sets and Systems, 55(1), 358--382,
2023.
[27] S. Opricovic and G. H. Tzeng, “Extended VIKOR method in comparison with outranking
methods,” Eur. J. Oper. Res., vol. 178, no. 2, pp. 514–529, 2007, doi: 10.1016/j.ejor.2006.01.020.
[28] J. Shieh, H. Wu, and K. Huang, “Knowledge-Based Systems A DEMATEL method in identifying
key success factors of hospital service quality,” Knowledge-Based Syst., vol. 23, no. 3, pp. 277–282,
2010, doi: 10.1016/j.knosys.2010.01.013.
[29] B. Han, G. Buchanan, and D. McKay, Learning in the Panopticon: Examining the Potential Impacts
of AI Monitoring on Students, vol. 1, no. 1. Association for Computing Machinery, 2022.
[30] S. Thiebes, S. Lins, and A. Sunyaev, “Trustworthy artificial intelligence,” Electron. Mark., vol. 31,
no. 2, pp. 447–464, 2021, doi: 10.1007/s12525-020-00441-4.
[31] O. Chaika, “Молодь і ринок,” no. October, 2023, doi: 10.24919/2617-0825.6/214.2023.
[32] N. N. T. Vy, “AI Implementation in ODR: A Game-Changer or a Troublemaker of Data
Protection,” Vietnamese J. Leg. Sci., vol. 8, no. 1, pp. 1–24, 2023, doi: 10.2478/vjls-2023-0001.
[33] M. Korir, S. Slade, W. Holmes, Y. Héliot, and B. Rienties, “Investigating the dimensions of
students’ privacy concern in the collection, use and sharing of data for learning analytics,” Comput.
Hum. Behav. Reports, vol. 9, p. 100262, 2023, doi: 10.1016/j.chbr.2022.100262.
[34] C. Mutimukwe, O. Viberg, L. M. Oberg, and T. Cerratto-Pargman, “Students’ privacy concerns in
learning analytics: Model development,” Br. J. Educ. Technol., vol. 53, no. 4, pp. 932–951, 2022, doi:
10.1111/bjet.13234.

40
 [35] J. P. McNulty and Y. Politis, “Empathy, emotional intelligence and interprofessional skills in
healthcare education,” J. Med. Imaging Radiat. Sci., vol. 54, no. 2, pp. 238–246, 2023, doi:
10.1016/j.jmir.2023.02.014.
[36] A. R. Pratama, F. M. Firmansyah, and F. Rahma, “Security awareness of single sign-on account in
the academic community: the roles of demographics, privacy concerns, and Big-Five personality,” PeerJ
Comput. Sci., vol. 8, pp. 1–20, 2022, doi: 10.7717/PEERJ-CS.918.
[37] F. Al-Sharqi, A. Al-Quran and Z. M. Rodzi, Multi-Attribute Group Decision-Making Based on
Aggregation Operator and Score Function of Bipolar Neutrosophic Hypersoft Environment,
Neutrosophic Sets and Systems, 61(1), 465-492, 2023.

41

View publication stats