Higher education embraces AI, yet neglects ethics and non-STEM fields

A major bibliometric study has spotlighted the rapidly evolving landscape of artificial intelligence (AI) applications in higher education, revealing critical insights into its current usage, gaps in implementation, and future research priorities. The research, titled “Using Artificial Intelligence for Higher Education: An Overview and Future Research Avenues”, was published in the open-access journal SAGE Open in May 2025.

Analyzing 181 peer-reviewed articles from Web of Science and Scopus databases, the study deploys co-occurrence techniques using VOSviewer and SciMat software to decode the conceptual structure and thematic evolution of AI in academia. It identifies the dominant themes, technological trends, and research clusters shaping the discourse, from AI’s role in academic performance and student engagement to its underutilization in management education and ethical oversight.

How is AI being used in higher education today?

The study finds that AI in higher education is predominantly associated with improving academic performance, online learning systems, and student retention. Chatbots like ChatGPT, plagiarism detection tools such as Turnitin, and adaptive platforms including Google Classroom are among the most widely discussed technologies. AI is being used to tailor learning pathways, offer real-time feedback, analyze student data for at-risk identification, and automate grading systems.

Cluster analysis reveals a strong focus on self-regulated learning, motivation, and student engagement. Emerging technologies like natural language processing and machine learning are also being employed to develop predictive models for educational success. Additionally, institutions are leveraging AI for language translation, educational content creation, and even robotics-enhanced learning in technical fields.

Despite this progress, the study warns that applications are heavily concentrated in STEM and language learning domains. There remains limited adoption in management and business studies, where the need for personalized study guides and streamlined academic administration could be well-served by AI technologies.

What are the emerging trends and research gaps?

The research outlines a significant evolution in themes between two periods, 1984 to 2020 and 2021 to 2023. The early period focused on foundational topics like robotics, programming, and curriculum support. In contrast, the recent surge in publications revolves around student performance, AI acceptance, gamification through chatbots, and e-learning frameworks. Concepts like engagement strategies and self-regulated learning have emerged as dominant clusters, particularly post-COVID-19, where hybrid learning reshaped institutional priorities.

Interestingly, the study finds that ethical issues, although recognized, remain underrepresented in academic literature. There is sparse discussion on the societal, legal, and moral ramifications of AI deployment in classrooms. Similarly, personalization, arguably one of AI’s most promising capabilities, is scarcely addressed in relation to student learning styles, institutional adaptability, or cultural context.

Student feedback and professor preparedness are also flagged as underexplored domains. While AI tools proliferate, the level of training provided to educators and the extent to which student voices shape tool adoption are minimal. This disconnect could limit the effectiveness and acceptance of AI-based systems in educational ecosystems.

Where should future research and policy focus?

The paper concludes by offering a comprehensive research agenda for closing existing gaps. It proposes empirical investigations into how chatbots like ChatGPT impact academic performance, how feedback mechanisms can be improved, and how engagement and self-regulated learning can be sustainably enhanced. It recommends targeted studies on AI applications in non-STEM fields like tourism, hospitality, and business education.

A core recommendation is to explore the potential of AI for educational equity through language-independent learning platforms and mobile education. Tools like real-time translators and AI tutors could enable students across linguistic and geographical boundaries to access higher education, promoting inclusion and diversity.

The authors also stress the importance of institutional strategy. Universities must assess how AI can support administrative tasks, reduce workload, and personalize curricula. Policymakers, too, are urged to invest in faculty development, ethical governance, and cross-sector partnerships to responsibly scale AI in education.

According to the analysis, most publications have emerged from technology-focused journals, with notable contributions from Spanish and Chinese researchers. However, there is a pressing need for broader international collaboration and a shift toward multidisciplinary inquiry, integrating perspectives from philosophy, law, and educational psychology into AI policy and practice.