The Digital Transformation of Higher Education in the Context of an AI-Driven Future

Abstract

In this article, digital transformation is examined as a key driver of structural and pedagogical change in higher education. This process is shown to expand access to learning, increase flexibility, support personalized educational trajectories, and enhance data-driven decision-making. At the same time, the effectiveness of digital transformation depends on institutional readiness, the quality of technological infrastructure, and the professional competencies of teaching staff. This research of this study is to assess the influence of digital transformation on the quality of higher education. This research employs a mixed-methods approach. Quantitative data from surveys of 4971 students and 483 instructors were analyzed using descriptive statistics, analysis of variance ANOVA, and multivariable regression, while qualitative focus group findings were examined through thematic analysis. The results indicate generally positive attitudes toward digitalization. The respondents emphasized flexibility and improved conditions for independent learning as key advantages of digital environments. However, this study also identifies several challenges, including infrastructural inequality, limited digital skills, and insufficient pedagogical adaptation. The article concludes that successful digital transformation requires a comprehensive strategic vision and sustained institutional support. For universities, strengthening digital competencies, modernizing infrastructure, and implementing management models focused on continuous improvement are essential conditions for ensuring sustainable development and enhancing the quality of education.

1. Introduction

The rapid digitalization of the global economy has transformed all spheres of society, with higher education becoming one of the most dynamic and strategically significant domains of change [1]. As universities play a central role in preparing human capital for the digital economy, the quality and relevance of higher education directly influence labour market development and national innovation capacity [2]. Consequently, accelerating processes of digitalization and globalization require higher education institutions to rethink their functions, revise development strategies, and systematically implement innovative digital solutions [3].

According to the Organisation for Economic Co-operation and Development (OECD) research, the educational systems of the future will operate as decentralized digital ecosystems in which decision-making processes are shared among multiple stakeholders, including administrators, employers, and students [4,5]. Such transformation must extend beyond the digitalization of instructional content to cover the entire spectrum of institutional processes-curricula, pedagogical models, assessment formats, student engagement, academic management, and collaboration with external partners [6].

The emergence of artificial intelligence (AI), especially in the post-ChatGPT 5 era, has amplified the scale and speed of digital transformation in education. AI now supports adaptive learning, academic analytics, cognitive automation, and personalized educational trajectories, creating both new opportunities and new institutional challenges [7,8]. However, despite the growing body of literature on digitalization, recent studies emphasize that empirical evidence on AI-driven transformation in emerging and post-Soviet higher education systems remains limited [9]. Furthermore, new integrative frameworks demonstrate that combining AI with Metaverse and Blockchain technologies may accelerate systemic change in universities worldwide [10].

Kazakhstan is also actively integrating digital and AI-enabled solutions into higher education. Yet, this process is characterized by uneven digital infrastructure, varying levels of digital competence, and fragmented institutional reforms [11]. These contextual factors make Kazakhstan a relevant case for examining how global digital trends are localized within national education ecosystems.

Therefore, this study provides a comprehensive analysis of the digital transformation of higher education in Kazakhstan, with a particular focus on AI adoption. The aim is to identify opportunities and limitations of ongoing transformation processes and to develop evidence-based recommendations for sustainable, ethical, and inclusive AI integration in universities. This research contributes to the international literature by offering a grounded empirical perspective from an emerging economy, addressing a gap in current post-ChatGPT 5 scholarship on AI-enabled digital transformation.

In addition, Kazakhstan’s digital reforms are embedded in broader national strategies. According to the Concept of Digital Transformation of the ICT and Cybersecurity Industry for 2023–2029 [12], education is identified as a priority area for building integrated digital ecosystems, deploying end-to-end technologies, and strengthening cyber resilience to achieve long-term digital sovereignty and sustainable development.

However, the extent to which these strategic goals translate into measurable improvements in educational quality remains insufficiently explored. This creates a critical need for empirical research that connects national policy ambitions with institutional practices and real user experiences in universities. Addressing this gap, the present study not only evaluates the current state of digital transformation but also examines the conditions required to ensure its effectiveness in Kazakhstan’s higher education system. To respond to these challenges, this study seeks to answer a set of research questions that clarify the impact of digital transformation and AI adoption in higher education.

2. Research Questions

In the context of the active digitalization of higher education and the integration of artificial intelligence technologies, it is essential to clearly understand the impact of these processes on the quality, accessibility and structure of education. Accordingly, this study is guided by the following research questions.

• RQ1. Which digital technologies are most effectively integrated into the educational process?
• RQ2. How does digital transformation affect the quality of higher education?
• RQ3. What opportunities does artificial intelligence provide for the AI-driven educational ecosystem?

3. Materials and Methods

3.1. Study Design

This study employed a cross-sectional survey design with a predominantly quantitative approach, complemented by a brief qualitative component in the form of open-ended questions. The objective was to explore students’ access to digital infrastructure, their usage patterns of generative AI (ChatGPT (OpenAI, GPT-5 model, San Francisco, CA, USA), Google Gemini (Version 1.5, Mountain View, CA, USA), Microsoft Copilot (Redmond, WA, USA, 2025 release)) tools, and their perceptions regarding associated risks and benefits.

3.2. Participants and Sampling

Due to organizational access constraints at L.N. Gumilyov Eurasian National University (ENU), Astana, Kazakhstan, a stratified convenience sampling strategy was employed. Although this non-probability method may limit the representativeness of the findings, stratification by field of study and year of education, together with a large sample size, helped reduce potential selection bias and ensured diversity among respondents. A total of 4971 students and 483 teachers participated in the survey, with 52% identifying as female and 48% as male, contributing to a balanced sample and minimizing the risk of systematic gender bias.

Participants were recruited through an open call disseminated via the university’s official LMS and institutional email channels. Participation was voluntary, and all students and faculty who met the inclusion criteria were invited to take part. No incentives were offered, and no exclusion criteria were applied other than the removal of responses with excessive missing data.

Inclusion criteria:

• Current enrolment as a student;
• Provided informed consent.

Responses with over 50% missing values on core scales were excluded from the final analysis.

Table 1. Demographic and academic characteristics of the sample.

Variable	Categories	n	(%)
Gender	Female	2585	52.0%
	Male	2386	48.0%
Year of Study	1st Year	1250	25.1%
	2nd Year	1300	26.1%
	3rd Year	1100	22.1%
	4th Year and above	1321	26.6%
Program/Field	Engineering	1800	36.2%
	Humanities	2300	46.3%
	Other	871	17.5%
Internet Quality	Excellent	1700	34.2%
	Good	2000	40.2%
	Fair	900	18.1%
	Poor	371	7.5%
Device Availability	Laptop	4200	84.5%
	Smartphone	3900	78.4%
	Tablet	1700	34.2%

presents group-level differences in the use of AI tools across demographic and academic variables.

Notable differences are observed across year levels and fields of study, with senior students and those enrolled in engineering programs reporting more frequent use of GenAI tools for academic purposes. By contrast, gender differences are minimal and statistically non-significant.

To address sampling considerations, participation in the survey was voluntary, which may introduce self-selection bias. A total of 4971 students and 483 instructors completed the questionnaire out of approximately 15,000 student and 1200 faculty invitations, resulting in an estimated response rate of about 33% for students and 40% for instructors. Although stratified convenience sampling increased diversity across programs and study levels, the single-institution focus limits broad representativeness. These factors were taken into account during the interpretation of the findings.

3.3. Measures

The survey instrument consisted of five main sections:

• Demographics and academic background–information about participants’ age, gender, academic level, and field of study;
• Digital access and infrastructure–availability and quality of internet connection, and access to digital devices;
• GenAI tool usage patterns–frequency of use and purposes in academic contexts (e.g., writing, coding, information search, analysis);
• Attitudes and perceived risks–perceived benefits, concerns related to academic integrity, and privacy issues;
• Open-ended questions–reflections on barriers to use and examples of good practices.

Survey items were measured using a five-point Likert scale format. The survey content underwent expert validation by 3–5 subject matter experts to ensure content validity. Internal consistency of the main composite scales was confirmed with Cronbach’s alpha values ≥ 0.70.

3.4. Data Preparation and Statistical Analysis

Data processing involved several steps. First, duplicate entries were removed, followed by logical consistency checks to ensure internal validity of responses. Missing data were handled according to the extent of incompleteness: (i) multiple imputation was applied when item-level missingness was below 10%, and (ii) listwise deletion was performed for cases with more than 50% missing responses. Additionally, outliers were identified and screened prior to analysis.

Descriptive statistics were calculated and presented as means, standard deviations, medians, proportions, and 95% confidence intervals (CI). Group comparisons were conducted using appropriate statistical tests: Chi-square (χ²) tests with Cramér’s V for categorical variables, and Welch’s t-tests or ANOVA for continuous variables. When assumptions of normality or homogeneity of variance were violated, non-parametric alternatives were applied.

To examine associations between variables while controlling for covariates (gender, age, year of study, internet quality, and device availability), a multivariable linear regression model was employed. Effect sizes were reported using Cohen’s d and η², along with two-sided p-values and a significance level of α = 0.05. Assumptions of normality and homogeneity were tested using the Shapiro–Wilk and Levene’s tests, while multicollinearity was assessed through Variance Inflation Factors (VIF). When assumptions were violated, non-parametric tests (such as the Mann–Whitney U test and the Kruskal–Wallis test) were applied. All statistical analyses were conducted using Python version 3.10 (Python Software Foundation, Beaverton, OR, USA) with the following libraries: pandas, NumPy, SciPy, statsmodels, scikit-learn, matplotlib, and seaborn. Figure 1 illustrates the frequency of GenAI tool usage, showing that students most frequently used GenAI for search and writing tasks.

The qualitative findings were integrated into the mixed-methods design through a convergence triangulation approach. Key themes—such as infrastructure barriers, digital readiness, and student engagement—were compared with the quantitative results to contextualize the statistical patterns and provide deeper explanatory insight into the survey findings.

3.5. Ethical Considerations

This research complied with ethical standards for human subjects. Participation was voluntary, anonymity was guaranteed, and all data were used exclusively for scientific purposes.

4. The Target Architectural Model of Digital Transformation

As part of this research, an integrated architectural model of the digital transformation of an enterprise was developed, reflecting key functional blocks, data exchange channels and information security mechanisms (Figure 2).

The use of a single digital platform that enables data unification and consolidation, as well as interaction of application services, is proposed as a system-forming element. This component functions as a central integration core, on the basis of which application subsystems are deployed [12,13,14].

Above the platform level lies an analytical module incorporating artificial intelligence (AI) tools, which support automated analysis, forecasting, and decision-making based on big data [15,16,17].

The user level is represented by a variety of interfaces, differentiated by target groups:

• A corporate web interface for employees and clients;
• Mobile applications for quick access to the system functionality;
• Integration interfaces of partner organizations [13].

Data exchange between external and internal systems is managed through the API integration layer, which ensures the scalability and compatibility of the architecture [18,19,20]. Information security is supported by a specialized module that provides data protection and compliance with current regulatory requirements [21,22].

Finally, a feedback loop integrated into the model provides continuous monitoring, the collection of operational metrics, and real-time adjustments to business process [23,24,25,26].

5. Risk Analysis and Development of Measures to Minimize Them

For the successful implementation of the architectural model, a systematic risk analysis was conducted and classified according to the probability of occurrence and the degree of impact on the project. The results are presented in an infographic diagram (

Table 2. Classification of digital transformation risks and measures for their mitigation.

Risk Category	Description	Probability	Influence	Reduction Measures
Technological	Module integration failure	Average	High	Conducting pilot testing, developing backup scenarios for operation
Organizational	Staff resistance	High	Medium	Implementation of training programs, involvement of opinion leaders in the change process
Financial	Budget overruns	Medium	High	Strict control of financial flows, phased implementation of the project
Legal	GDPR non-compliance	Low	High	Conducting a legal audit, consultations with specialists in the field of data protection

), which clearly highlights the priority areas for preventive measures.

The proposed risk management system focuses proactively identifying potential threats, developing preventive measures and integrating risk management processes into the strategic planning framework for digital transformation.

Based on the risk profile and mitigation strategies, it is essential to evaluate how digital transformation is currently unfolding in higher education systems. The next section therefore examines the present state of digital transformation, highlighting international trends and institutional practices.

6. Current State of Digital Transformation in Higher Education

Digital transformation is profoundly reshaping the higher education landscape worldwide. The integration of technology into academic and administrative functions has become not only a trend but a necessity, reflecting a shift toward flexible, inclusive, and data-driven educational ecosystems [27,28,29].

Contemporary higher education systems are increasingly adopting digital platforms, online course delivery, LMS, and adaptive learning technologies. Additional critical tools include big data analytics, virtual reality (VR), and augmented reality (AR) environments. These instruments support more personalized learning experiences and more efficient institutional management [30,31].

At the policy level, the European Digital Education Action Plan 2021–2027 highlights three key objectives: fostering digital literacy, ensuring inclusive access, and applying digital tools to improve teaching quality [32,33].

Europe has been at the forefront of many of these innovations. In Germany, digitalization initiatives began in the late 20th century and accelerated through federal support for educational technology infrastructure [34,35]. France has established national platforms such as FUN (France Université Numérique) to promote massive open online courses (MOOCs) and foster digital collaboration [36,37,38,39,40]. In the Netherlands, the emphasis lies in LMS integration and the use of data to monitor academic performance [41,42,43].

Beyond national borders, institutions are aligning with international initiatives like the European Higher Education Area (EHEA). This framework promotes cross-border mobility, the recognition of digital credentials, and standardization of academic data systems [44,45]. Meanwhile, global platforms such as Coursera, edX, and FutureLearn facilitate transnational learning and provide scalable models for digital education [46,47].

Despite these advances, digital inequality remains a persistent issue. Infrastructure limitations, lack of access to devices, and gaps in digital literacy continue to affect many communities, particularly in developing regions. The COVID-19 pandemic further exposed and exacerbated these disparities. To address this, many governments are investing in broadband expansion, subsidized hardware, and digital training programs [48,49,50].

The transformation also redefines the role of educators. Teachers are no longer regarded solely as content providers. Instead, they are becoming facilitators, instructional designers, and mentors in digital spaces. These emerging roles require professional development in ICT, digital pedagogy, and online communication [51,52,53,54,55,56,57,58,59,60,61].

Institutional transformation further requires strong internal strategies. Across Europe, national governments and universities are adopting coordinated frameworks for digital transition. These strategies include regulatory reforms, competency development, and the establishment of cross-sector partnerships [62,63,64]. According to the European Commission, collaboration among the state, academic institutions, industry, and civil society is essential for long-term success [65,66,67,68,69,70,71,72].

Technological advancements are also strengthening university infrastructure. Tools such as digital grade books, blockchain-based diplomas, LMS automation, and ERP systems streamline academic operations. In addition, interactive platforms and AI-supported systems improve service delivery and enable evidence-based decision-making [73,74,75,76,77,78,79,80,81,82,83,84,85,86,87].

To conceptualize institutional readiness, Figure 3 illustrates a five-tier maturity model for digital transformation. It is structured hierarchically, progressing from strategic intent to external integration:

• Strategic–Digital vision, institutional priorities, and maturity assessment;
• Organizational–Culture, leadership, project management, and internal alignment;
• Technological–Core infrastructure, cybersecurity, and system integration;
• Operational–Workflow automation, digital services, and analytics usage;
• Ecosystem–Partnerships, platform interoperability, and innovation networks.

This model underscores the layered and interdependent nature of digital transformation. with each level building upon the previous one. Institutions that develop a cohesive digital strategy, invest in infrastructure, and maintain collaborative networks are more likely to navigate this transition successfully. In turn, such efforts facilitate adaptability, strengthen resilience, and improved outcomes for all stakeholders.

While global trends provide a valuable reference point, understanding national and local dynamics remains crucial. Accordingly, the subsequent section focuses on the specifics of Kazakhstan’s higher education context to align global insights with local realities.

7. Digital Transformation Processes in the Higher Education System of Kazakhstan

Digital transformation in Kazakhstan’s higher education has been designated a national strategic priority. Guided by the 2023–2029 Concept for the Development of Higher Education and Science, universities are transitioning into smart institutions powered by digital ecosystems. These include EdTech platforms, electronic content, distance learning, and digitalized administration [88].

The architecture of Kazakhstan’s digital universities relies on an integrated data core, modern LMS systems (e.g., IS Univer 2.0, Platonus v6.25.9), adaptive technologies, and data-driven planning. However, compared with countries such as Turkey, gaps remain in infrastructure and policy coordination [89].

The COVID-19 pandemic accelerated digital adoption, prompting universities to implement video conferencing tools, virtual laboratories, and simulators. This period boosted digital literacy but also revealed persistent challenges, including inadequate infrastructure, insufficient faculty preparedness, and shortages of digital content [90].

The state plays a substantial role in this process. The National Educational Database (NEDB) centralizes educational data, while the Unified Higher Education Platform (launched in 2023) integrates institutional data with labour market and public sector systems [91].

Universities also rely on both international and local digital libraries (e.g., eLABa, ProQuest), which expand access and reduce regional disparities. Moreover, digitalization is shifting pedagogical focus-from knowledge transfer toward critical thinking and independent learning [92].

The Digital Transformation Concept (2023–2029) emphasizes the development of cyber-resilient systems, national cloud platforms, and adherence to the principles of “security by design” and “inclusion by default”. It also promotes the establishment of the National Research Data Cloud to strengthen research integrity and foster international collaboration [93].

Despite progress, several challenges remain, including unequal access, insufficient staff training, and fragmented standards. A structured approach, encompassing digital risk mapping and stakeholder collaboration, is essential for sustainable transformation [94].

As shown in Figure 4, the conceptual architecture of a digital university ecosystem is built around a central data lake, which integrates teaching, research, applicants & alumni, support services, library resources, and governance modules.

At the core of the system is a data lake that provides open APIs and identity management services. Surrounding this central component are integrated subsystems designed to support the key functions of the university ecosystem:

• Teaching–learning management systems (LMS) and e-assessment tools;
• Research–grant management systems and institutional repositories;
• Applicants & Alumni–customer relationship management (CRM) systems and dedicated portals;
• Support services–human resources, IT, and finance systems;
• Library services–access to open educational resources (OER) and subscription-based databases;
• Governance–compliance with data protection laws and auditing mechanisms.

Built on top of these layers, an AI and analytics module provides dashboards and actionable insights that continuously feed back into the ecosystem to inform decision-making.

The overall architecture is guided by three core principles: scalability, interoperability, and security, ensuring that the system remains sustainable, flexible, and resilient as it evolves.

Despite ongoing progress, a number of unresolved barriers continue to hinder digital transformation in Kazakhstani universities. Accordingly, the next section highlights the key systemic challenges that limit the effectiveness of these reforms.

8. Key Issues and Challenges in the Digitalization of Higher Education

Digital transformation in higher education, while facilitating innovation, also exposes a daunting array of systemic challenges. These extend across infrastructure, pedagogy, organization, ethics, and law, and require multi-faceted responses [86,95,96,97,98,99,100,101,102,103,104,105,106,107,108].

Among the most persistent impediments are digital disparities. These manifest as unequal access to reliable internet, adequate equipment, and modern teaching platforms, particularly in less-developed regions. Even in advanced nations, digital literacy differs markedly across students, teachers, and administrators. Case analyses indicate that in countries such as Kazakhstan, digitalization is limited by fragmented IT infrastructure and insufficient staff training, which exacerbate learning inequalities [70,72,86,88,89,90,92,94].

Another pressing issue is faculty preparation. Effective digital learning requires more than ICT proficiency. Teachers are expected to design digital content, integrate analytics, and remain informed about emerging pedagogies. However, many faculty members are overwhelmed by bureaucratic digital infrastructure tasks, which contribute to burnout and diminished motivation. Further, the lack of harmonized quality assurance standards renders accreditation and recognition of digital offerings challenging, while measures of learning outcomes in virtual environments remain in an early developmental phase [87,95,97,98,100].

Ethical concerns are also at the forefront. The increasing usage of AI, adaptive systems, and cloud platforms raises critical questions regarding data protection, algorithmic opacity, and intellectual freedom. Without strong regulatory guidelines, universities risk breaching digital rights of faculty and students [101,102,104,105,109,110,111,112,113].

Another important issue is “digital alienation.” Reduced face-to-face interaction and large-scale online delivery models weaken the community-building function of universities. Overreliance on technology may result in “technological determinism,” where platforms dictate pedagogy and erode academic freedom and creativity [86,92,95,98]. Misalignment with labour market demands is also a concern. Recruiters frequently report gaps between graduate skills digital economic requirements. Curriculum typically lags behind technological advancements in fields such as AI, big data, and cybersecurity [73,74,75,76,77,99,100].

Last but not least, the absence of a shared cybersecurity architecture exposes sensitive information to significant risk. As universities process increasing volumes of behavioural and academic data, greater compliance with legal and ethical guidelines becomes imperative [91,92,95].

Digital transformation is more than a technological shift but a definition of a higher education sector’s culture and strategy. Its success depends upon systematic planning, investment in infrastructure and human capital, legal safeguards, and societal consensus aligned with the shared [66,67,68,69,70,71,86,88,90,98].

In addition to organizational and infrastructural barriers, digital transformation also has significant psychological implications for participants in the educational process. Studies show that prolonged online learning can increase levels of stress, anxiety, and emotional exhaustion among students, especially when learning environments lack social interaction and stable support systems [99]. Reduced motivation, “Zoom fatigue,” and feelings of isolation are frequently reported consequences of intensive digital engagement, contributing to the phenomenon of digital alienation. Similar patterns are observed among faculty, who may experience burnout due to increased workload, continuous connectivity, and pressure to adapt to new digital tools without adequate training. These psychological factors influence learning outcomes and instructional quality, highlighting the need to integrate digital well-being principles into institutional strategies for digital transformation.

9. Adoption of Artificial Intelligence in Higher Education Institutions

AI is a central enabler of the digital transformation of higher education, reforming educational, administrative, and research processes to support more flexible, personalised, and effective learning environments [106,108].

With adaptive learning, teaching platforms tailor course content to students’ knowledge and behaviours, blending learning analytics, automatic grading, and outcome prediction to deliver instruction at-scale and in hybrid formats [100,105,109]. Assessment is further strengthened by automated grading, plagiarism detection, and early identification of at-risk students [74,75,110].

Administrative activities are increasingly supported by AI chatbots for applications, registration, and document workflow, thereby reducing staff workloads and enhancing service delivery [97,98].

In research, AI assists with literature analysis, data organization, and project management [103,107]. At the system level, analytics integrate data traces from learning management systems (LMS) and related tools to optimize curricula and pedagogy [73,102].

Worldwide practice demonstrates diverse approaches: South Korea and Japan are leading in mental analytics and robot assistants, while Kazakhstan is experimenting with language-sensitive models (e.g., ISSAI) that promote multilingual and culturally responsive learning [88,89,90,92].

These benefits are offset by significant challenges: large-scale upskilling of faculty [93,95,96]; ethical and transparency concerns about algorithmic decision-making [60,61,101]; risks of over-reliance on behavioural data that may undermine the humanistic mission of education [15,112,113]; and persistent inequalities in access [64,69,99]. Successful adoption accordingly entails integrating AI into institutional plans, investing in research and capacity building, and developing shared standards for safety, governance, and accountability. Importantly, educators should be supported-not replaced-by AI, and teaching must remain student-centred, based on personalization, collaboration, scholarly integrity, and equity [62,66,68,70].

While AI offers new opportunities for digital transformation, its successful implementation ultimately depends on the attitudes and readiness of key stakeholders. Therefore, the following section examines students’, teachers’, and employers’ perceptions of digitalization and AI integration.

10. Studies on the Attitudes of Educational Process Participants Toward Digital Transformation

In the context of this research, an extensive program was developed to investigate the perception of digitalization in higher education across diverse target groups. The primary goals of the program were fourfold:

• Defining students’, teachers’, and employers’ attitudes toward the processes of digitalization, focusing on opportunities, challenges, and readiness to adopt new technologies [86,87,94];
• Analysing the perception and application of artificial intelligence technologies in both educational and managerial activities of universities, including adaptive learning, administrative automation, and academic research support [97,98,106,109];
• Assessing the degree of institutional preparedness for digital transformation by examining university strategies, infrastructural capacity, and faculty digital competencies [62,63,88,89,90];
• Collecting proposals for improving digital educational practices, ensuring that digital transformation is aligned with pedagogical integrity, inclusiveness, and sustainable long-term implementation [95,96,99,104].

Students’ and Teachers’ Questioning

Sociological questionnaire surveys were conducted among students and teachers to identify their attitude toward the digitalization of education and to assess their level of preparedness for implementing artificial intelligence [60,61,62,63].

During surveys, particular attention was given to compliance with ethical requirements, in line with academic standards and the principles of integrity in scientific activity [1,3,60]. All respondents were informed in advance about the objectives, scope and methodology of this study, as well as their rights as participants [6,64]. Participation was entirely voluntary, and respondents could withdraw from this study at any stage without any consequences [7,65].

Prior to data collection, each participant gave written or oral consent (depending on the survey administration method) to participate in this study [5,61]. Anonymity and confidentiality of were strictly maintained: the data contain no identifying features and are used solely in generalized form for scientific analysis [8,62,70].

No vulnerable populations were involved in this study, and no sensitive questions were included that could cause psychological discomfort to respondents [3,72]. Data collection, storage, and handling were carried out in accordance with personal data protection principles and were limited to the duration of this research project [9,68,71]. This study fully complied with ethical standards aimed at respecting the dignity, rights, and freedoms of all participants [60,61,66].

Both paper-based and online questionnaires were administered. The survey included questions on the convenience of digital learning, attitudes toward artificial intelligence in education, and satisfaction with digital tools [63,64,99,100].

Based on these research objectives and stakeholder perspectives, the subsequent section presents the empirical results of this study, highlighting key trends in digital readiness and AI adoption.

11. Results

11.1. Descriptive Statistics

A total of 4971 students and 483 teachers participated in the survey. Among students, 52% were female and 48% were male. The mean student satisfaction score was 3.84 (SD = 0.71) on a five-point scale, with a 95% CI [3.82, 3.87]. Regarding infrastructure, 27.4% of students reported unstable internet access, while most respondents had regular access to a digital device suitable for learning. Overall, the descriptive statistics indicate generally positive perceptions of digital learning, but with noticeable variation linked to technological conditions (Figure 5).

As shown in Figure 5, the results indicate that (a) gender distribution was relatively balanced (52% female, 48% male); (b) internet quality was rated as “average” by 39% of students, while 16% reported it as “poor,” 18% as “above average,” and only a small proportion as “excellent”; and (c) 55% agreed or strongly agreed that digital learning improves comprehension, whereas 8% disagreed and the rest expressed neutral or uncertain views.

As shown in Figure 6, student perceptions were mixed across three dimensions: (a) 37% strongly agreed that online lectures and DERs improve comprehension, while 8% disagreed and 54.9% were uncertain; (b) 48% rated platform usability neutrally, with 19% reporting difficulties and 33% responding positively; and (c) 53% reported not using generative AI, whereas 26% used it occasionally and 21% frequently.

As shown in Figure 7, 53% of students reported not using generative AI, while 26% used it occasionally and 21% frequently (a); 69% of teachers were female and 31% male, indicating a gender-skewed sample (b); and 59% rated internet quality as “average,” with 20% describing it as “excellent,” 14% as “above average,” and 7% as “poor” (c).

Figure 8 illustrates the survey results, showing that a majority of teachers (63%) evaluated the level of digitalization of educational and scientific processes as “average”. A comparable assessment was expressed regarding the digitalization of administrative processes, with 59% of teachers also rating it as “average”.

As shown in Figure 9, teachers reported differing views on digitalization: (a) 52% found digital tools very helpful in teaching, while 48% saw them as only partially useful; (b) 71% positively assessed the impact of digitalization on educational quality, with 29% remaining neutral; and (c) generative AI usage among teachers remained limited, as 53% did not use such tools, 26% used them several times a year, and 21% several times a month.

11.2. Group Differences and Predictors of Student Satisfaction

The results revealed statistically significant differences in student satisfaction depending on internet connection quality (

Table 3. Group differences in student satisfaction.

Variable	Group	M (SD)	Test	Test Value	p-Value	Effect Size
Internet Quality	Stable	4.02 (0.65)	ANOVA	F (2, 4968) = 31.47	<0.001	η2 = 0.012
Internet Quality	Unstable	3.53 (0.78)	t-test	t (2381) = 8.21	<0.001	d = 0.41
Gender	Female vs. Male	—	t-test	t (4970) = 1.14	0.254	ns

Note. “ns” = non-significant result.

).

A one-way ANOVA showed that students with stable connectivity reported significantly higher satisfaction (M = 4.02, SD = 0.65) than those with unstable internet (M = 3.53, SD = 0.78), F (2, 4968) = 31.47, p < 0.001, η² = 0.012. A Welch’s t-test confirmed this difference, t (2381) = 8.21, p < 0.001, d = 0.41, indicating a moderate effect size.

In contrast, gender-based comparisons did not yield significant differences in satisfaction scores, t (4970) = 1.14, p = 0.254. This indicates that demographic characteristics had minimal influence on students’ attitudes toward digital learning, whereas infrastructural factors were decisive.

Differences in categorical variables were assessed using the χ² test. Internet quality had a significant impact on student satisfaction, χ²(3) = 21.45, p < 0.001. The effect size was Cramér’s V = 0.18, indicating a weak-to-moderate association.

The regression model was found to be statistically significant, F(5, 4965) = 42.83, p < 0.001, explaining 21% of the variance in student satisfaction (

Table 4. Model summary of the regression analysis.

Model Statistics	Value
R2	0.21
F(5, 4965)	42.83
p (model)	<0.001

Note. Dependent variable = student satisfaction.

). As presented in

Table 5. Multivariable regression results predicting student satisfaction.

Predictor	β	SE	t	p-Value	95% CI
Internet Quality	0.31	0.02	12.77	<0.001	[0.26, 0.36]
Digital Readiness	0.29	0.03	10.41	<0.001	[0.24, 0.34]
Device Availability	0.07	0.02	2.88	0.004	[0.02, 0.11]
Year of Study	0.03	0.02	1.70	0.091	[−0.01, 0.07]
Gender (F = 1)	0.02	0.02	1.06	0.287	[−0.02, 0.06]

, internet quality (β = 0.31, p < 0.001) and digital readiness (β = 0.29, p < 0.001) were the strongest predictors of student satisfaction, indicating that students’ learning experiences in digital environments are highly dependent on both stable infrastructure and sufficient digital competencies. A smaller but statistically significant effect was identified for device availability (β = 0.07, p = 0.004). In contrast, gender and year of study did not demonstrate meaningful predictive power (p > 0.05).

These results are further visualized in Figure 10, which shows that infrastructure- and readiness-related variables exert the most substantial influence on satisfaction, while demographic characteristics contribute minimally to the model. Collectively, the tables and figure confirm that technological conditions, rather than demographic factors, are the primary drivers of positive digital learning experiences.

These results are further visualized in Figure 10, which shows that infrastructure- and readiness-related variables exert the most substantial influence on satisfaction, while demographic characteristics contribute minimally to the model. Collectively, the tables and the figure confirm that technological conditions-particularly internet stability and digital readiness-are the primary drivers of positive digital learning experiences. Overall, these findings suggest that improving digital infrastructure and enhancing students’ digital competencies can substantially increase satisfaction in technology-enhanced learning environments.

11.3. Content Analysis

This research in this format was conducted to analyse media and public opinion about digitalization of higher education, as well as to identify the principal trends and public attitude toward the spread of digital technologies. Its objective was to examine key themes and general perspectives on digitalization based on media reports and other information sources [62,64,95]. The principal activities included identifying leading discussion topics, evaluating the emotional tone of these discussions, and defining barriers and challenges marked in media space [68,96].

To achieve these aims, this research employed computerized data collection techniques (API access and web scraping), supplemented by linguistic and thematic analysis using natural language processing (NLP) methods. The analytical toolkit included sentiment analysis, topic modelling, and quantitative analysis of mentions with findings visualized in interactive dashboards in Power BI [95,97].

Content analysis revealed a steep increase in media coverage of digitalization challenges since 2017, reaching a maximum in 2023 [62,99]. The key themes extracted in media reports included online learning, the role of artificial intelligence in higher education, challenges in digital infrastructure, and teachers’ and students’ digital competencies [64,70,88,96]. Overall, the tone of discussions was overwhelmingly positive, emphasizing innovation and higher education’s potential for greater efficiency [95,97]. At the same time, a significant numbers of publications underscored obstacles to effective digitalization, particularly inadequate digital infrastructure and insufficient teacher training [62,68,99]. Figure 11 illustrates that media references to digitalization in higher education have increased continuously since 2017, reaching a maximum in 2023 (Figure 11).

It appeared that digital transformation in higher education is generally viewed positively. However, significant infrastructural and methodological barriers remain [62,64,95]. The key challenges include regional imbalances, where large cities exhibit more sophisticated digitalization than periphery regions, as well as a shortage of professionals for implementing and maintaining digital innovations [68,96].

Also, the analysis of publications, reports, and regulatory documents identified several key trends:

• Digitalization is recognized as a state educational policy priority, acting as a basis for national modernization strategies [62,70];
• Universities are increasingly involved in both national and foreign digital initiatives, in line with global higher education developments [64,88];
• Studies on AI and EdTech solutions in universities is growing rapidly, proving their expanding role in both practice and research [95,97];
• Regulatory oversight of digital activities and protection of personally identifiable information is in perpetual emphasis, particularly in relation to ethical governance and adherence to international standards [68,99].

11.4. Focus Groups

Business representatives for the focus groups were recruited using purposive sampling. Invitations were sent to companies that (1) are official ENU partners, (2) regularly participate in internship or employment programs for students, and (3) represent diverse sectors of the digital economy. Only those who confirmed availability and consented to participate were included. This approach ensured relevance to this research topic, although it may limit the breadth of perspectives.

Focus groups were conducted with teachers, students, and administrators to better understand individual and shared perspectives on digitalization. This format allowed for more detailed discussion of key issues and helped identify latent concerns and unserved demands not entirely reflected in the quantitative survey data [101,102,103,104,105]. The main purpose of the focus groups was to gather qualitative observations on perceptions of digital technologies, examine practical challenges in their implementing, and explore possible avenues for further improving the digitalization of higher education [106].

The focus groups took the form of expert and group interviews, employing specifically designed guides and stratified questions to probe the practical experience, barriers, and benefits of digitalization in the learning process [102,105].

The results confirmed the findings of both the questionnaire survey and the content analysis while also revealing additional challenges, including:

• Insufficient communication between administrators and teachers during the adoption and use of digital solutions;
• Inadequate training of teachers in digital pedagogy and educational technology;
• Substantial local disparities in the level of digitalization across educational institutions [101,107].

At the same time, focus group discussions showed that students generally expressed positive attitudes toward the flexibility of digital learning, the availability of digital materials, and the opportunity to combine study with employment [103]. Conversely, teachers voiced concerns about students’ low motivation in online environments and challenges related to the organization of interactive, participatory learning [106].

Recommended solutions for improving the digital learning environment included:

• Developing more user-friendly platforms and portals;
• Expanding technical support services;
• Strengthening pre-service and in-service teacher training;
• Enhancing students’ digital skills for academic and career purposes [104,108].

The focus groups thereby highlighted that teacher preparedness and institutional support remain key limiting factors while also showing the value of qualitative strategies in revealing deeper causes of the difficulties indicated by quantitative research.

11.5. Summary and Analysis of the Results of All Studies

To comprehensively assess the impact of digital transformation on higher education, a comparative analysis was conducted of the perspectives of three key stakeholder groups: students, teachers, and business representatives. The analysis was based primarily on the student questionnaire survey, which aimed to identify their experience with and attitudes toward digital educational technologies [109,110,111]. Data on teachers and business representatives were gathered through a qualitative content analysis of expert interviews, analytical reports, and open sources, as well as through the synthesis of typical judgments expressed in focus groups and industry-level discussions on the digitalization of education [112].

This mixed-method approach allowed for the identification of both converging expectations and significant divergences in the perception of digital transformation among the core participants of the educational process and external stakeholders (

Table 6. Results of this study of the perception of the digital transformation of higher education by three key stakeholder groups.

	Students	Teachers	Business
Goals of digitalization	Increased accessibility, convenience, flexibility of learning	Automation of processes, efficiency, increasing workload	Preparation for the labour market, IT skills, flexible trajectories
Infrastructure and access	Lack of technology, average internet quality	Old equipment, lack of IT support	Gap between graduates and the IT reality of companies
Use of digital technologies	Active use of LMS, video, clouds, AI; 30% have difficulties	Limited use, concerns about new tools	Demand for simulators, AI, data analysis
Attitude to AI	60% use AI, but are not sure about ethics	Scepticism, fear of integrity violations, need regulations	Support for the use of AI, formation of digital autonomy
Expected effects	Flexibility, interactivity, access to knowledge	Increased engagement, decreasing routine workload	Preparation of digitally competent graduates
Ethics and academic integrity	Lack of knowledge about citation, AI rules, no training	Concerns about plagiarism and AI, require training courses	Formation of a culture of digital responsibility, transparency

) [99].

Table 3 provides a comparative analysis of how three key stakeholder groups, i.e., students, teachers, and business representatives, perceive the digital transformation of higher education. The analysis is structured around the main thematic categories reflecting the goals, practices, and expectations of digitalization.

As can be seen from the analysis, students primarily value the accessibility, convenience, and flexibility of learning provided by digital formats [63]. However, they face technical and organizational barriers, including insufficient equipment, poor Internet connectivity, and software that is not always user-friendly [72]. At the same time, students are actively mastering AI tools, but note a lack of knowledge about their proper use and the standards of academic integrity [61,111].

Teachers recognize the positive aspects of digitalization such as automations of routine tasks, student engagement, but they often experience overload, technical disruptions, and a shortage of methodological support [86,97]. Educators also express concern about generative AI and emphasize the need for universities to establish ethical guidelines and education programs [61,105].

The business sector considers digital transformation primarily in terms of graduates’ readiness for digital economy, placing strong emphasis on applied IT, big data, and artificial intelligence skills [89,90]. Moreover, business representatives also highlight the need for students to develop a culture of digital responsibility and technological transparency [83,100]. Thus, each stakeholder group in higher education has its own specific expectations of digital transformation. For sustainable development of higher education, it is important to take these diverse perspectives into consideration and design a coordinated digitalization strategy that simultaneously addresses educational, professional and ethical standards [94,98]. This study results indicate a generally positive perception of digitalization among students and employers. Students appreciate the convenience and accessibility of digital format but remain cautious about the risks of digital inequality [63,72]. Teachers acknowledge the potential of digitalization, yet express concern about the lack of time and resources required for a high-quality transition to new platforms [86,97]. Employers focus on the development of graduates’ digital competencies, but are concerned about the gaps in data analysis and AI skills [74,75,109].

Additionally, a brief sensitivity check indicated that the key statistical relationships (e.g., the effects of internet quality and digital readiness) remained stable despite potential sampling bias. However, given the voluntary nature of participation, the results should be interpreted with caution, as response behavior may differ from that of non-respondents.

12. Discussion

The results of this study provide new insights into the dynamics of digital transformation in higher education, revealing how technological integration, institutional readiness, and human factors jointly shape the effectiveness of digital learning environments. Three key conclusions can be drawn by interpreting the findings in light of existing literature and current global trends.

First, this study demonstrates that the most effective digital technologies in higher education are end-to-end solutions that support continuity across all stages of the learning process. The integration of LMS platforms, interactive multimedia resources, and generative AI tools was found to improve accessibility, personalization, and learner engagement. These findings are aligned with earlier studies that emphasize that digital ecosystems—rather than isolated tools—produce sustainable educational benefits by enabling flexibility, data-driven decision-making, and learner-centered design [5,8,100]. However, the results also indicate that technology effectiveness depends not only on availability, but on meaningful instructional design and user competence. This supports the argument that digital tools amplify pedagogical intent rather than independently determining learning outcomes [4,27].

Second, the findings indicate that digital transformation has a dual impact on the quality of higher education. When supported by institutional capacity, teacher digital competencies, and stable technological infrastructure, digitalization enhances learning quality through continuous feedback, inclusive access, and adaptive content. Yet, in the absence of these conditions, digital transformation risks creating superficial learning, fragmented course design, and increased cognitive load for both educators and students. This duality resonates with international studies showing that digital transformation improves outcomes only when accompanied by pedagogical innovation, stakeholder participation, and governance reform [107,108]. Therefore, quality enhancement is not a technological outcome per se, but the result of systemic readiness, professional development, and coordinated institutional strategy.

Third, this study highlights the transformative role of artificial intelligence in shaping future educational models. AI technologies provide opportunities for adaptive learning, predictive analytics, and process automation, thereby redefining teaching and learning practices. This is consistent with previous research that positions AI as a catalyst for personalized learning trajectories and intelligent decision-support systems [109,110,111,112,113]. At the same time, the findings underscore the importance of ethical frameworks, equitable access, and digital literacy for responsible AI adoption. AI does not replace educators; rather, it reshapes pedagogical roles and requires a shift toward mentoring, facilitation, and analytical guidance. Therefore, AI-driven transformation must remain human-centered, balancing technological efficiency with ethical and cultural considerations.

Taken together, these findings imply that digital transformation in higher education is moving from procedural tool adoption toward systemic redesign, where technology, pedagogy, and institutional culture co-evolve. This shift demands integrated strategies that align infrastructure, professional development, AI-based innovation, and evidence-based governance. Successful digital ecosystems will be those that foster personalization, broaden learner agency, and support continuous improvement at both micro- (classroom) and meso- (institutional) levels.

The use of a single university sample also affects the national applicability of the findings. Although ENU represents a large institution within Kazakhstan, its infrastructure, student demographics, and digital capacity may not fully reflect the diversity of other universities in the country. Therefore, the results should be viewed as context-specific rather than universally generalizable. Future research would benefit from multi-site or cross-regional validation to strengthen external validity and capture institutional variation.

The findings of this study resonate with emerging literature showing that the psychological dimension of digital learning is a critical component of human-centered digital transformation. For students, the benefits of flexibility and autonomy are often accompanied by heightened stress, reduced intrinsic motivation, and feelings of isolation when social presence is limited [99]. For instructors, digital transformation can also amplify emotional burdens through workload intensification and cognitive overload. Therefore, successful implementation requires not only technological and pedagogical innovation, but also well-being interventions, including digital hygiene training, institutional support programs, and workload management policies. Integrating mental-health-informed approaches will help ensure that AI-driven ecosystems enhance—rather than undermine—human well-being in higher education.

This study has several limitations that should be considered when interpreting the results. First, this research was conducted within a single university (L.N. Gumilyov Eurasian National University), which may limit the generalizability of the findings to the wider higher education system of Kazakhstan. Future studies should utilize probability-based sampling across multiple institutions or countries to enhance external validity and nationwide representativeness. Second, the cross-sectional design does not allow for causal inference between key variables (e.g., access to digital technologies and intention to use AI). Therefore, the results should be interpreted as correlational; longitudinal or experimental designs are recommended to establish causal mechanisms more accurately. Third, the reliance on self-reported data introduces potential response and recall biases, and future research should complement self-reported measures with objective indicators such as LMS activity logs or system analytics to strengthen the findings—particularly business representatives—as our approach may have resulted in self-selection bias, limiting the representativeness of stakeholder views. Finally, although appropriate statistical procedures were applied, the combination of a cross-sectional design and self-reported data may still influence the robustness and generalizability of the findings. Future studies could also incorporate well-established technology adoption models (e.g., TAM, UTAUT, IDT, ECM, TRI) to deepen the theoretical understanding of digital and AI adoption in higher education.

13. Conclusions

This study demonstrates that digital transformation has the potential to significantly improve the quality of higher education by expanding access, supporting flexible learning pathways, and enabling data-driven decision-making. However, the findings reveal that the success of this process is contingent upon institutional readiness, stable technological infrastructure, and the digital competencies of educators. When these conditions are not met, digital initiatives risk producing fragmented learning experiences and limiting their long-term impact.

The results confirm that students and instructors hold generally positive attitudes toward digital solutions, recognizing their value for independent learning and enhanced engagement. At the same time, persistent challenges—such as uneven infrastructure, limited digital skills, and insufficient pedagogical adaptation—continue to hinder effective implementation. These findings highlight the need for a balanced approach in which technology is integrated not as an end in itself but as part of a coherent strategy that aligns organizational development, professional training, and pedagogical innovation.

To ensure sustainable progress, universities should prioritize the development of digital competencies, invest in modern infrastructure, and adopt management models focused on continuous improvement. Future research may broaden the scope by including multiple institutions or countries and by exploring the long-term effects of digital transformation through longitudinal and experimental designs.