Exploring the impact of artificial intelligence on higher education: The dynamics of ethical, social, and educational implications

Artificial Intelligence (AI) is reshaping multiple sectors, including education, through advances that enable personalization, automation, analytics, and efficiency gains. In higher education, AI tools such as adaptive learning platforms, chatbots, automated grading, and data analytics are increasingly present, offering benefits but also raising concerns around data privacy and security, algorithmic bias, fairness, transparency, and accountability. As AI gains prominence, there is a need to explore its impact on ethical, social, and educational dynamics in higher education, and to understand how different stakeholders (students, faculty, administrators) perceive AI and envision its future. This study addresses key objectives: (1) examine attitudes and perceptions toward AI in higher education, including concerns and expectations; (2) assess AI’s impact on teaching and learning (benefits and drawbacks for instruction, personalization, and outcomes); (3) investigate ethical and social implications (privacy, bias, fairness, transparency, accountability); (4) explore the envisioned future role of AI in higher education; and (5) analyze how demographic characteristics shape perspectives on AI’s ethical, social, and educational dynamics. Research questions: RQ1: What are participants’ attitudes and perceptions towards the implementation of AI in higher education? RQ2: What is the role of AI in teaching and learning in higher education? RQ3: What ethical and social implications arise from the implementation of AI in higher education? RQ4: How do participants envision the future role of AI in higher education? RQ5: How do participants’ demographic characteristics impact their perspectives in terms of the ethical, social, and educational dynamics associated with AI implementation?

The literature on AI in higher education has expanded rapidly, spanning education, computer science, psychology, and ethics. Key domains include: (1) Pedagogical innovations: AI (e.g., intelligent tutoring, adaptive platforms) supports personalized learning, student engagement, and improved outcomes; it can transform teaching, assessment, access, retention, costs, and administration, with evidence of positive impacts from chatbots and tutoring systems. Identified gaps include limited focus on higher-order thinking, collaboration/communication, AI skills, and large-scale implementation integrated with instructional practice. (2) Learning analytics and student support: AI-driven analytics use student data (including affect) to identify at-risk learners, recommend interventions, and provide timely feedback. Studies demonstrate effective prediction, optimized grouping, and analysis of collaboration patterns. Gaps include ethical considerations, transparency, and accountability in analytics use. (3) Assessment and grading: AI (NLP, plagiarism detection, automated scoring) can reduce workload and enable data-driven decisions with reliability and validity comparable to traditional grading. The literature urges attention to practical impact on learning and institutional support for analytics-driven environments. (4) Educators’ professional development: AI can enhance instructors’ adaptive strategies and attitudes; institutions should address training and support needs. Examples include ML-driven analysis of student feedback to inform teaching improvement. (5) Ethical and social implications: Core concerns include privacy/security, bias/fairness, transparency, accountability, preserving human roles, and societal impacts (inequality, workforce shifts). Tabled gaps highlight the need for comprehensive exploration of ethical concerns, societal impacts, and workforce transformation alongside technical integration.

Design: Quantitative cross-sectional study using an online survey questionnaire targeting higher education stakeholders in Saudi Arabia. Sample: N = 1,113 participants comprising students (83.6%), faculty (10.1%), and administrators (6.3%). Demographics: Age—24 or less (77.9%), 25–34 (9.4%), 35–44 (8.6%), 45+ (4.0); Gender—Male (44.7%), Female (55.3%); Education level—Bachelor (85.0%), Master (6.2%), PhD (8.8%); Majors—Medicine/Engineering/Computer Science (63.8%), Literary/Humanities/Education (21.7%), Business/Commerce/Law (14.5%); Subjective AI expertise—Low (46.5%), Medium (43.9%), High (9.6%); Usage frequency—Daily (32.8%), Weekly (21.1%), Rarely/Monthly/infrequent reported as per figures. Instrument: Survey measured four subscales—Attitudes and Perceptions; Role of AI in Teaching and Learning; Ethical and Social Implications; Future Role of AI. Reliability: Cronbach’s alpha—Attitudes and Perceptions α=0.92; Role of AI in Teaching and Learning α=0.92; Ethical and Social Implications α=0.89; Future Role of AI α=0.93; Total α=0.96. Additional items captured AI tools/services used, purposes of usage, and negative experiences. Analysis: Descriptive statistics (means, SDs) for all items and composite scales; Multivariate Analyses of Variance (MANOVAs) examined effects of total AI uses, purposes, and difficulties on the dependent composites (attitudes/perceptions; role in teaching/learning; ethical/social implications; future role). Multivariate effects: Uses—Wilks’ Λ=0.492, F=2.845, p<0.001, η²=0.163; Purposes—Λ=0.547, F=2.272, p<0.001, η²=0.140; Difficulties—Λ=0.614, F=1.898, p<0.001, η²=0.115. Follow-up tests of between-subjects effects indicated significant relationships between uses/purposes/difficulties and several outcome composites (details in Table 10). Context: Focus on higher education in Saudi Arabia; data availability noted as supplementary material.

- Sample: N=1,113; predominantly students (83.6%), age ≤24 (77.9%), female (55.3%). - AI tools/services (means): Face recognition 4.32 (SD 1.11), Speech recognition 3.92 (1.18), AI-chatting tools 3.85 (1.25), AI-powered design/creativity 3.60 (1.37). - Purposes of usage (means): General 4.38 (0.80), Educational 4.35 (0.90), Research 4.18 (1.06), Entertainment 3.96 (1.07), e-Government 3.49 (1.38), Commercial 3.34 (1.36), Google AI services 3.60 (1.30). - Negative experiences (means): Privacy/security 3.37 (1.25), Technical issues during installation 3.35 (1.27), Technical issues during usage 3.15 (1.16), Usage difficulties 3.02 (1.27). - Attitudes and perceptions (Table 5): Strongly positive overall (Total M=4.33, SD=0.71). Highest-rated items: enhancing learning experience (M=4.43), improving access to resources (M=4.42), improving student outcomes (M=4.34). - Role of AI in teaching and learning (Table 6): Positive overall (Total M=4.21, SD=0.75). Highlights: improved accessibility (M=4.31), personalized learning (M=4.30), automating administrative tasks (M=4.27), adaptive environments (M=4.26). - Ethical and social implications (Table 7): Strong agreement (Total M=4.37, SD=0.64). Highest items: establish ethical guidelines (M=4.47), respect student autonomy (M=4.45), avoid exacerbating inequalities (M=4.42), retain human support (M=4.40), address bias/fairness (M=4.37). - Future role of AI (Table 8): Positive expectations (Total M=4.30, SD=0.73). Top items: intelligent tutoring systems (M=4.36), prioritizing ethics/human values (M=4.35), transforming teaching/learning (M=4.35). - MANOVA results (Table 9): Significant multivariate effects of total uses (Λ=0.492, F=2.845, p<0.001, η²=0.163), total purposes (Λ=0.547, F=2.272, p<0.001, η²=0.140), and total difficulties (Λ=0.614, F=1.898, p<0.001, η²=0.115) on outcome variables, indicating that how often and for what purposes AI is used, as well as encountered difficulties, significantly shape attitudes, perceptions, and future expectations. - Between-subjects tests (Table 10): Significant associations reported between uses/purposes and all four composites; difficulties showed significant effects on Role of AI in Teaching and Learning and Future role composites.

Findings address the research questions by demonstrating broad, positive stakeholder attitudes toward AI’s integration in higher education (RQ1) and recognition of concrete roles for AI in teaching and learning (RQ2), including personalization, accessibility, administrative efficiency, adaptive environments, and real-time performance insights. Participants simultaneously emphasized ethical and social imperatives (RQ3), calling for governance frameworks that ensure privacy, security, fairness, transparency, accountability, respect for student autonomy, and preservation of human support. Regarding the future (RQ4), stakeholders anticipate intelligent tutoring, ethically guided integration, transformed pedagogy, personalized learning pathways, enhanced assessment, and support for lifelong learning. Multivariate analyses (RQ5) show that AI usage patterns, purposes, and experienced difficulties significantly affect stakeholders’ attitudes and expectations, underscoring the importance of practical exposure and intentionality in AI implementation. Collectively, the results suggest institutions should pair technical deployment with robust ethical governance, capacity building, and support structures to maximize benefits while mitigating risks.

The study evidences widespread optimism among higher education stakeholders in Saudi Arabia about AI’s capacity to enhance teaching and learning, resource access, administrative processes, and institutional innovation. Participants favorably rate common AI tools (face/speech recognition, chatbots) while urging progress in more advanced applications. Ethical governance emerges as essential, with priority on privacy, security, bias mitigation, transparency, accountability, and preserving human roles. Practical experience with AI, clarity of purpose, and addressing implementation challenges shape positive attitudes and future expectations. Implications for policy and practice include: (1) develop and implement ethical guidelines for AI integration; (2) invest in professional development and training for faculty, administrators, and students; (3) provide resources and infrastructure to support AI; (4) encourage collaboration and interdisciplinary research; (5) address data ethics and privacy; (6) establish evaluation frameworks for AI impact; (7) foster industry partnerships; and (8) continuously monitor and adapt AI implementations. Future work should broaden contexts and comparisons to refine generalizability and deepen understanding of socio-technical factors.

- Reliance on self-reported data may introduce bias and inaccuracies relative to actual behaviors or attitudes. - Limited consideration of contextual factors (cultural, institutional, regional) that may shape perceptions of AI. - Focus confined to higher education; broader societal implications and perspectives from other sectors are not examined. Future research should conduct cross-cultural and comparative studies, interdisciplinary research, and comparative analyses with other educational contexts to capture contextual influences and broader societal implications.