AI Adoption in Higher Education: A Comparative Study of Institutional Readiness and Challenges
Open accessAbstract
This study explored the patterns of AI adoption, varying degrees of institutional readiness, and challenges of implementation in higher education in Pakistan, employing a mixed-methods comparative approach. To collect relevant data, the researchers used structured questionnaires from 250 faculty members and administrators and conducted semi-structured interviews with 30 key stakeholders in 5 sampled universities, including public, private, and federally chartered institutions. The results indicated stark differences in the level of AI readiness from different institution types; specifically, private universities showed more advanced infrastructural capability (3.8/5.0), as opposed to public (2.9) and federal institutions (3.2). Notably, while 72% of respondents reported a general awareness of AI, only 22% of institutions were reported to have AI applications, which were primarily operated as pilots and not fully integrated into the institution. The predominant barrier to implementation was identified as a lack of financial resources (82%), followed by insufficient infrastructure (78%), and a lack of technical expertise (74%). From the qualitative assessment, five thematic challenges of implementation emerged: gaps in leadership vision, deficits in technical capacity, faculty resistance to change, financial sustainability concerns, and absence of policy frameworks. Additionally, the lack of comprehensive change management was evident, as 64% of respondents expressed concerns about job security and 72% about data privacy. These findings suggest the need for phased approaches, along with substantial infrastructural investments, training, and clear policy frameworks to address the AI adoption imbalance in higher education institutions in Pakistan.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
1 of 7 cited sources checked · 0 supported · 0 flagged · accuracy not yet measured
- Findingown data
General awareness of AI was reported by 72% of the overall respondent group.
“while 72% of respondents reported a general awareness of AI”
- Also stated in: § Methodology · plain
“General AI Awareness 68% 82% 65% 72%”
Quote in the paperNot a substantial claim - Also stated in: § Methodology · plain
- Findingown data
Substantial AI understanding was present in just 38% of the sample.
“Substantial understanding was found in just 38% of respondents”
Quote in the paperNot a substantial claim - Findingown data
The top barrier was financial constraints, rated major or critical by 82% of respondents.
“The predominant barrier to implementation was identified as a lack of financial resources (82%)”
- Also stated in: § Methodology · plain
“Financial constraints emerged as the most critical challenge (82% rated it major/critical)”
Quote in the paperSubstantial claim - Also stated in: § Methodology · plain
- Findingown data
Most AI implementations were pilot or limited-scale projects rather than fully integrated institution-wide deployments.
“which were primarily operated as pilots and not fully integrated into the institution.”
- Also stated in: § Methodology · plain
“Most implementations, at 68%, stayed at the pilot project stage instead of becoming institutionwide deployments.”
Quote in the paperSubstantial claim - Also stated in: § Methodology · plain
- Findingown data
Private universities had 38% current AI systems, compared with 12% in public universities.
“private universities led with 38% compared to just 12% in public ones.”
Quote in the paperSubstantial claim
- Findingown data
Younger faculty were more positive about AI (78%) than senior colleagues (42%).
“younger faculty members (78% positive) were more receptive, in contrast to just 42% positivity among their senior colleagues.”
Quote in the paperNot a substantial claim - Findingown data
Comprehensive operational plans for AI integration or adoption were absent at most institutions.
“Most institutions lack comprehensive operational AI integration or adoption plans, forcing leaders to improvise.”
Quote in the paperSubstantial claim - Findingown data
Data privacy concerns were expressed by 72% of respondents.
“and 72% about data privacy.”
- Also stated in: § Methodology · plain
“72% cited data privacy fears.”
Quote in the paperSubstantial claim - Also stated in: § Methodology · plain
- Findingown data
AI applications were present in just 22% of institutions.
“only 22% of institutions were reported to have AI applications”
- Also stated in: § Methodology · plain
“Currently Have AI Systems 12% 38% 22% 22%”
Quote in the paperSubstantial claim - Also stated in: § Methodology · plain
- Findingown data
Formal AI training had reached only 24% of respondents.
“Formal AI training reached only 24%”
Quote in the paperSubstantial claim - Findingown data
Favorable views of AI integration correlated strongly with previous technology exposure.
“those with prior technology exposure showed a strong correlation with favorable perceptions of AI integration.”
Quote in the paperNot a substantial claim - Findingown data
A major stakeholder theme was the lack of articulated institution-wide AI policies.
“A major theme raised by stakeholders was the absence of articulated, institution-wide policies for AI adoption.”
Quote in the paperSubstantial claim - Findingown data
70% of respondents agreed that AI would improve education quality.
“Although 70% agreed AI would improve education quality”
Quote in the paperNot a substantial claim - Findingown data
Institutions' hesitation to fully adopt AI was mainly attributed to technical and infrastructure limitations.
“Technical challenges and infrastructure limitations were identified as the main reasons for institutions' hesitance to fully adopt AI.”
Quote in the paperNot a substantial claim - Findingown data
Private universities had the highest good/excellent internet connectivity (78%), versus public (45%) and federal (62%).
“Good/Excellent Internet Connectivity 45% 78% 62% 62%”
Quote in the paperNot a substantial claim - Findingown data
Infrastructure inadequacy was a barrier for 78% of institutions.
“followed by insufficient infrastructure (78%)”
- Also stated in: § Methodology · plain
“Infrastructure inadequacy affected 78% of institutions”
Quote in the paperSubstantial claim - Also stated in: § Methodology · plain
- Findingown data
Insufficient technical expertise was cited by 74%.
“and a lack of technical expertise (74%).”
- Also stated in: § Methodology · plain
“while 74% cited insufficient technical expertise.”
Quote in the paperSubstantial claim - Also stated in: § Methodology · plain
- Findingnone given
Financial difficulties affected all institutions, irrespective of their type.
“Every institution faces financial challenges, regardless of its type.”
Quote in the paperNot a substantial claim - Findingown data
Pilot-stage projects accounted for 68% of implementations, not institution-wide rollouts.
“Most implementations, at 68%, stayed at the pilot project stage instead of becoming institutionwide deployments.”
Quote in the paperSubstantial claim - Findingown data
65% of respondents could identify AI applications.
“however, 65% could identify AI applications.”
Quote in the paperNot a substantial claim - Findingown data
Job security concerns were expressed by 64% of respondents.
“64% of respondents expressed concerns about job security”
- Also stated in: § Methodology · plain
“64% had job security worries”
Quote in the paperSubstantial claim - Also stated in: § Methodology · plain
- Findingown data
Private universities had a higher mean infrastructure readiness score (3.8 out of 5.0) than public (2.9) and federal (3.2) institutions.
“private universities showed more advanced infrastructural capability (3.8/5.0), as opposed to public (2.9) and federal institutions (3.2).”
- Also stated in: § Methodology · plain
“Private universities led in readiness (3.8/5.0),ahead of public (2.9) and federal (3.2) institutions.”
Quote in the paperSubstantial claim - Also stated in: § Methodology · plain
- Interpretationown data
Greater readiness and adoption at private universities, compared with public and federal institutions, was mainly attributed to infrastructure and financial resource gaps.
“Private universities demonstrated greater readiness and adoption than public and federally chartered institutions.This was mainly due to differences in infrastructure and financial resources.”
Quote in the paperSubstantial claim- Rests on an assumption: The five sampled universities represent public, private, and federally chartered higher education institutions in Pakistan.
- Interpretationown data
Professional training fell short of the organizational capability improvement and expansion that was needed.
“The professional training offered did not align with the necessary improvement and expansion of organizational capabilities.”
Quote in the paperSubstantial claim - Premisecitation
Focused studies on AI adoption in South Asian higher education are clearly lacking.
“there is a clear lack of focused studies on artificial intelligence adoption in South Asian higher education (Henadirage and Gunarathne 2025).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Henadirage and Gunarathne 2025 source
- Interpretationown data
The comparative framework indicated that readiness extended beyond technical factors to leadership, governance, strategic planning, change management, and stakeholder inclusion.
“The comparative framework showed institutional readiness involved more than technical factors.It included leadership, governance, strategic planning, change management protocols, and stakeholder inclusion.”
Quote in the paperSubstantial claim - Interpretationown data
Frameworks for technology developed in high-resource settings need substantial adaptation for developing countries.
“the study showed that technology frameworks developed in high-resource contexts need considerable adjustments to work in the developing world.”
Quote in the paperSubstantial claim - Premisecitation
Effective policies and resource allocation were made difficult by the lack of comprehensive research on institutional AI preparedness.
“The lack of comprehensive studies on institutional preparedness for AI integration made it difficult to shape effective policies and allocate resources (Aslam et al., 2025).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Aslam et al., 2025
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- May combine two claims. 13
- Classified as probably not a substantial claim. 5
May combine two claims.
PremisecitationEven advanced technologies faced adoption barriers in Pakistani universities due to institutional inertia and insufficient training, as seen with e-learning and LMS.
“The experience with e-learning and learning management systems in Pakistani universities demonstrates that even advanced technologies face adoption barriers due to institutional inertia and insufficient training (Ahmad, Mohd Noor et al. 2023).”
Quote in the paperNot a substantial claim1 of 1 sources checked- cites Ahmad, Mohd Noor et al. 2023 source
The Computer Self-Efficacy section (p. 8) describes resistance to LMS adoption, inadequate instructor training, and a lack of coordinated guidance from university administrators. The Professional Training and Network Infrastructure section (p. 8) also identifies training as a barrier. These points are discussed across higher education generally, however; the paper’s Pakistan-specific example concerns unreliable and costly internet access (Technology Factor, p. 10), not institutional inertia or insufficient training in Pakistani universities. (LLM, conf=0.87)
Experimental — accuracy not yet measured
Probably not a substantial claim: may not pass S3 (load-bearing).
Findingown dataThe HEC's missing national policy guidance was identified by stakeholders as a cause of inconsistent institutional policies.
“Stakeholders identified the Higher Education Commission's lack of national policy guidance as a reason other institutions display inconsistent policies.”
Quote in the paperNot a substantial claimMay combine two claims.
Findingown dataDefinite future AI plans were reported by 65% of private universities and 35% of public universities.
“65% of private universities are committed to definite future AI plans, far surpassing the 35% of public institutions”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataPrivate universities reported 85% LMS adoption, while public universities reported 52%.
“Learning Management System adoption was 85% in private versus 52% in public universities”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataChange resistance was notable at 68%, particularly among senior faculty.
“Resistance to change was significant (68%), especially among senior faculty.”
Quote in the paperNot a substantial claimMay combine two claims.
Findingown dataQualitative analysis produced five implementation themes: leadership vision gaps, technical capacity deficits, faculty resistance, financial sustainability concerns, and missing policy frameworks.
“five thematic challenges of implementation emerged: gaps in leadership vision, deficits in technical capacity, faculty resistance to change, financial sustainability concerns, and absence of policy frameworks.”
- Also stated in: § Qualitative Analysis Thematic Insights from Stakeholder Interviews · plain
“Analysis of the qualitative data identified five themes.”
Quote in the paperSubstantial claim- Also stated in: § Qualitative Analysis Thematic Insights from Stakeholder Interviews · plain
May combine two claims.
Findingown dataAdoption was substantially hindered by data security worries and missing policy.
“Data security concerns and policy absence also hindered adoption efforts substantially.”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataIT received 12% of the budget in private universities and 4.5% in public universities.
“They allocated 12% of their budget to IT, while public universities allocated 4.5%.”
Quote in the paperNot a substantial claimMay combine two claims.
Interpretationown dataTighter budgets and legacy systems were among the more serious obstacles facing public institutions.
“In contrast, public institutions encounter more serious obstacles, such as tighter budgets and legacy systems that do not embrace modern technology.”
Quote in the paperSubstantial claimMay combine two claims.
Interpretationown dataPrivate higher education institutions showed more advanced AI adoption, attributed mainly to superior technology, greater financial capacity, and more flexible administration.
“The study shows AI adoption is more advanced in private higher education institutions, primarily due to more sophisticated technology, better disposable income, and more flexible administration.”
Quote in the paperSubstantial claim- Rests on an assumption: The five sampled universities represent public, private, and federally chartered higher education institutions in Pakistan.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationBetter institutional readiness leads to improved technology implementation, adoption, and return on investment.
“Higher institutional readiness improves technology implementation, adoption, and return on investment (Wong and Li, 2025).”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Wong and Li, 2025 source
Probably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductown dataThe study's findings were strengthened by consistency between its quantitative and qualitative data.
“Alignment between quantitative and qualitative data strengthened the findings of the study.”
Quote in the paperNot a substantial claimMay combine two claims.
PremisecitationPakistani challenges such as adoption resistance, insufficient professional training, and a policy-practice gap mirror wider educational technology findings.
“In Pakistan, challenges such as resistance to adoption, lack of professional training, and a gap between policy and practice reflect the broader issues identified in educational technology research (Qazi, Sharif et al. 2024).”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Qazi, Sharif et al. 2024 source
May combine two claims.
PremisecitationAdoption of AI in Pakistan's higher education is uneven and shows gaps in preparedness, implementation, and outcomes.
“However, adoption is inconsistent, revealing gaps in preparedness, implementation, and outcomes (Aslam et al., 2025).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Aslam et al., 2025
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Study design and conductnone givenEquitable distribution was ensured through purposive sampling.
“Purposive sampling ensured equitable distribution.”
Quote in the paperNot a substantial claimMay combine two claims.
PremisecitationInfrastructural problems, uneven digital literacy among staff and students, and constrained technology budgets characterize Pakistani universities.
“Pakistani universities operate within a context of infrastructural challenges, uneven digital literacy among teachers and students, and limited budgets for technology (Ikram and Kenayathulla et al., 2025).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Ikram and Kenayathulla et al., 2025 source
Probably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductnone givenThe study satisfied ethical requirements by obtaining informed consent and institutional approvals.
“Ethical requirements were met through informed consent and institutional approvals.”
Quote in the paperNot a substantial claim
The paper relies on this without support.
The five sampled universities represent public, private, and federally chartered higher education institutions in Pakistan.
“conducting a comparative case study of five universities from different regions.The sample included public, private, and federally chartered institutions.”
The paper relies on this without support.
Questionnaire responses accurately reflect respondents' actual AI awareness, readiness, and attitudes.
“Two hundred fifty faculty members and administrators completed structured questionnaires, assessing institutional readiness, technological infrastructure, and AI adoption factors.”
The paper relies on this without support.
The Technology Acceptance Model and Diffusion of Innovation Theory are suitable frameworks for assessing AI readiness in Pakistani higher education.
“The study used the Technology Acceptance Model and Diffusion of Innovation Theory to guide the readiness framework.”
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Discussion reverses the reported deployment pattern
MajorThe discussion says most deployments are institution-wide, whereas the results say most remain pilots. These are opposing descriptions of the same implementation pattern.
“Most deployments are institution-wide, not pilot projects, which points to sustainability and scalability as major challenges for first implementations.”
“Most implementations, at 68%, stayed at the pilot project stage instead of becoming institutionwide deployments.”
Reported share of universities cannot represent the five studied
MinorIf the study covered five universities, 22% of them would be 1.1 universities, not a possible count. The adoption percentage may describe respondent reports, but the quoted statement explicitly attributes it to universities.
“Only 22% of universities adopted AI systems;”
“conducting a comparative case study of five universities from different regions.”
Cites
The publisher hasn’t deposited this paper’s reference list.
Research works printed but not linked: 4
- [4]4. Aslam, M. W. (2025). A Comprehensive Review of Wearable Health Devices: Advances, Challenges, and Future Directions.
- [5]5. Aslam, M. W. (2025). Financial Feasibility of Implementing Smart Safety Technologies in Electrical Engineering Projects: A Review of Current Status and Future Prospects. Spectrum of Engineering Sciences, 3(3), 557-567.
- [6]6. Aslam, M. W. (2025). THE ROLE OF ARTIFICIAL INTELLIGENCE IN ELECTRICAL ENGINEERING APPLICATIONS IN SMART GRIDS, POWER SYSTEMS, AND AUTOMATION. Spectrum of Engineering Sciences, 3(3), 540-556.
- [7]7. Chu, H. C. (2022). Roles and research trends of artificial intelligence in higher education: A systematic review of the top 50 most-cited articles. Australasian Journal of Educational Technology, 38(3), 22-42.
Cited by
Sorted by citations
Nyamboga, T. O. (2026) · Frontiers in Education · doi:10.3389/feduc.2026.1899369 · Cited by 1 · see citing works
Sutton, L. B.-M., & Degenaar, A. (2026) · Corporate Communications An International Journal · doi:10.1108/ccij-09-2025-0274
Ljepava, N., & Selaković, M. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/407424
Strategic Leadership for Building Responsible AI Ecosystems in Education
Salendab, F. A. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-2600-0467-8.ch002
The adaptive learning taxonomy for responsible large language model integration in higher education
Baradziej, S., & Kochanska, A. (2026) · Discover Artificial Intelligence · doi:10.1007/s44163-026-01528-1
“Huma et al. [18] highlight in their comparative study of institutional readiness that while technical infrastructure for AI often exists, the necessary ethical frameworks, faculty training, and policy guidelines are frequently absent or underdeveloped Kasneci et al. [19].”
Christauskas, Č., & Staugaitis, A. J. (2026) · Acta Scientiarum Polonorum - Oeconomia · doi:10.22630/aspe.2026.25.2.5
“From a managerial and organizational perspective, leadership, policy clarity, and investment in digital infrastructure are critical enablers of effective AI adoption in educational settings (Huma et al., 2025; Shahzad et al., 2025).”
AI Governance Maturity Framework Using Interpretable PLS-SEM and Machine Learning Analytics
Gambe, M. S., & Arcos, J. R. D. (2026) · International Conference on Innovations in Cybersecurity and Data Science (ICICDS) · doi:10.1109/icicds70526.2026.11604675
Khalid, A. B. E., & Akram, M. (2026) · Qlantic journal of social sciences. · doi:10.55737/10.55737/qjss.vii-ii.26510
Gender Difference in EPI (Expanded Program of Immunization) Awareness among Teacher Educators
Waqas, M., & Akram, M. (2026) · Qlantic journal of social sciences. · doi:10.55737/qjss.vii-iii.26556
- Authors
- Huma, T., Ahmed, D. S., Mahmood, W., khan, D. A., & Afridi, D. A. K.
- Affiliations
- Iqra University (PK)
- University of Balochistan (PK)
- INTI International University (MY)
- Universitat de Barcelona (ES)
- University of Peshawar (PK)
- Venue
- Social Science Review Archives
- Published
- 2025-10-13
- Volume
- 3
- Issue
- 4
- Pages
- 304-312
- ISSN
- 3006-4694 (ISSN-L)
- 3006-4708
- OpenAlex
- W7095008371
- Semantic Scholar
- 05ea636c87aa869819a757fa4df8a91da322bb43
- Cited by
- 9
- References
- 0
- Citations by year
- 2026: 9
- FWCI
- 11.17
- Type
- journal article
- Open access
- diamond
- doi.org
- Locations
- Social science review archives. (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Information Systems › Artificial Intelligence in Education
- Physical Sciences › Computer Science › Information Systems › Educational Innovations and Challenges
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence Applications
- Keywords
- artificial intelligence adoption, higher education, institutional readiness, implementation barriers, public and private universities, financial constraints, technical expertise, faculty resistance to change, leadership vision, AI policy frameworks, change management, data privacy, job security, Pakistan
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-09-28.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-09-28.
Cited sources are checked by gpt-6-sol.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
