Benefits and Risks of Artificial Intelligence in Higher Education


There are several benefits for professors, students, and institutions from implementing Artificial Intelligence (AI) in higher education. In general, universities have found that different AI tools have benefitted both educators and students (Pisica et al., 2023). These benefits include:

  • Grading – Educators in higher education benefit from AI systems designed to grade students’ assignments, quizzes, and exams, giving them valuable written feedback in the process. This reduces the amount of time spent on manually grading assessments and allows educators to keep track of students’ progress in an efficient way (Owoc et al., 2021; Chan & Hu, 2023).
  • Assessments – AI can generate exam questions and even tailor them to specific students to address areas of weakness. Professors benefit from the time saved on writing exam questions and students can focus on improving (Dave & Patel, 2023).
  • Feedback  Both students and professors benefit from AI-facilitated feedback. Chatbots allow students to give more relevant and valuable feedback to professors through interactive dialogue rather than through a survey. (Owoc et al., 2021).
  • Personalized Learning – To prevent students from falling behind in their courses, AI benefits students who need to learn at their own pace and have lessons tailored to their needs (Owoc et al., 2021).
  • Cheating – AI detection tools for cheating and plagiarism benefits educators and institutions striving for academic integrity (Owoc et al., 2021)
  • Administration – At the institutional level, AI is beneficial in supporting administrative efficiency, day-to-day functionality, and digital security (Pisica et al., 2023).

AI can be adapted to different fields of higher education to suit educators’ and students’ particular needs. In medical and dental education, students benefit from “virtual simulation and training” by applying their knowledge in a practical setting without the ethical concerns related to using it on real patients. AI may improve students’ critical thinking, problem-solving, and clinical reasoning skills: instead of exams that test memorization, AI may generate exams that emphasize these skills (Dave & Patel, 2023; Sallam et al., 2023). In second language learning, students may use chatbots to develop their writing skills and gain confidence in their command of the language. Non-native speakers benefit from the writing assistance provided by generative AI (Chan & Hu, 2023).

For researchers, certain AI tools are equipped to search for and find relevant sources and draw from methods from other disciplines to apply to their specific research projects (Pisica et al., 2023). AI helps them come up with ideas, synthesize information, and summarize large amounts of text, which facilitates the publication process. These tools help students focus on analysis rather than source-gathering (Chan & Hu, 2023).



One of the greatest risks associated with AI in higher education is, as anywhere else, data security being compromised. The “organization, collection, control, storage, usage, archival and destruction” of personal data of both students and educators require resources for those actions to be carried out responsibly. Students are concerned about personal data collection, which prevents them from trusting AI systems and risks them becoming averse to AI in higher education (Owoc et al., 2021; Chan & Hu, 2023; Sallam et al., 2023).

Impact on Transversal Skills

Specific to medical students, there is a risk of overreliance on AI that might impede the development of critical thinking and problem-solving skills at the diagnosis and treatment level. For exams, AI may also lack predictability, creativity, and “ethical considerations” when generating questions. It puts students at risk by being less challenged or over-challenged, thereby becoming demoralized and less motivated to learn (Dave & Patel, 2023). Overreliance on generative AI tools may also impact the development of students’ writing abilities, as well as intellectual development overall (Chan & Hu, 2023).

Impact on Emotional Intelligence and Communication

If universities do not aim to strive for a balance between traditional methods of education and AI-facilitated education, they risk reducing students’ exposure to physical interaction and environments where they develop their emotional and social intelligence skills. It may also affect professor-student relationships and create barriers for students in seeking individual help. In healthcare, students risk not developing the skills needed to provide patients with emotional support or build a healthy professional-client relationship (Pisica et al., 2023; Chan & Hu, 2023; Sallam et al., 2023). 

Bias and Discrimination

In healthcare education, the inaccuracy and biases of AI may lead to output that students take at face value but may run the risk of being unethical or discriminatory. It may reinforce misconceptions rather than help students reassess them (Sallam et al., 2023; Rasul et al.).

Compromised Academic Integrity

Another risk associated with AI, particularly ChatGPT, is students’ ability to cheat by generating their essays and assignments to artificially inflate their grades and hinder their ability to learn by not doing the work themselves. If not properly educated on the ethics of AI usage, students may risk missing out on assignments that aim to develop their “academic formation” (Dave & Patel, 2023; Pisica et al., 2023; Chan & Hu, 2023).

Digital Inequity

While AI implementation may benefit students in general, there is also the risk that it creates knowledge disparity between students who have access to “technology and high-speed internet” and those who do not. The responsibility falls on the academic institution to ensure students have equal access to such technology (Rasul et al.).

Incomplete Implementation

Universities may have issues with implementing AI due to poor infrastructure and a lack of technological knowledge among their users. This leads to the risk of incomplete or inefficient implementation of AI and providing an inadequate digital education to students. Additionally, upgrading software and incorporating constantly evolving AI technologies is costly, which runs the risk of universities either allocating funds they do not have or falling behind new advancements (Pisica et al., 2023; Rasul et al.).

Benefits and Risks of AI Implementation in Romania


Global Competitiveness

Romanian universities face pressure from various societal trends such as globalization, demographic challenges, and the shift towards a more technological society. They need to balance students’ needs and expectations while operating in a global, digital context. Implementing AI in higher education will allow Romania to remain competitive on a global scale (Grosseck, 2019).

Transformation of the Teaching-Learning Process

In Romania specifically, both professors and students have discussed how AI in Romanian universities can help provide “a better educational experience” by introducing “new learning content processes, and methods.” It allows professors to develop new curricula, use tools tailored to the needs of their specific courses, and introduce new forms of “teaching and assessment.” (Pisica et al., 2023). This includes “learning by doing,” problem-solving learning, and “developing transversal competencies and transferable knowledge” (Bucea-Manea-Toniș et al., 2022). Academics from Spiru Haret University have also discussed how personalized learning through virtual tutoring, adaptation to students’ needs, and creating new curricula benefit the teaching-learning process (Pisica et al., 2023).

AI has the potential to enhance professional success by improving efficiency in planning class content, saving valuable time. It can also elevate student performance and engagement by personalizing the learning process. These advancements significantly contribute to the overall quality of teaching. Teachers can leverage AI to select suitable materials for lectures, reduce errors, gain deeper insights into student performance, identify students who may be at risk, analyze incorrect answers, and provide additional instruction where needed. It may also be used to assess assignments and detect plagiarism (Bucea-Manea-Toniș et al., 2022).


Furthermore, the utilization of AI in research is one of the greater benefits. This is demonstrated by its support function in the research process, facilitating international collaboration with researchers from universities abroad, enhancing the peer review process, and encouraging “interdisciplinary, multidisciplinary, and transdisciplinary research” (Pisica et al., 2023).

Improvement of Transversal and Transferable Skills

Addressing the skills needed to succeed in the current and future job market, academics also note that AI and digitalization in higher education will benefit students through the skills they will develop in using such technologies. Through developing transversal and transferable skills, AI implementation will help Romanian students to prepare for future work and transition from university to the job market. It affords students the ability to “critically access, use, analyze and create digital content.” (Pisica et al., 2023; Bucea-Manea-Toniș et al., 2022; Grosseck, 2019).


Students in Romania will also benefit from the accessibility granted by AI implementation. With AI, higher education will be able to support students with different backgrounds and special needs more easily. Classrooms will benefit from more “student participation” and creativity (Pisica et al., 2023; Bucea-Manea-Toniș et al., 2022).


In terms of economics, institutions in Romania can benefit from reducing the cost of administrative tasks, as AI can carry those out. Students benefit from more easily accessible information and communication with university administrators (Pisica et al., 2023).


Impact on Skills Development and Communication

The risks associated with implementing AI in Romanian universities overlap with foreign universities. Such risks include using AI without knowing the impact it may have on students’ development of “critical thinking [skills], building and developing moral values, or sharing responsibility.” Reliance on AI may lead to cheating and negatively impact students’ skills development. Another concern is the socio-psychological effects of students spending more time in a “virtual environment” instead of interacting with their peers and professors. Without proper care, students can lose motivation and become unable to communicate with others, impacting professors’ motivation to teach as well (Pisica et al., 2023). 


Another risk relevant to Romania is compromised security, with concerns over data breaches and loss of privacy for professors and students being among the most common. “AI safety research” and new laws can minimize the risk of identity theft, bullying, and discrimination (Pisica et al., 2023; Bucea-Manea-Toniș et al., 2022).

Incomplete Implementation

The incomplete implementation also applies to Romanian universities. Institutions may implement AI but then risk professors not being trained to understand and implement AI appropriately, as well as not having adequate staff with relevant knowledge to maintain AI systems. The costliness of incomplete AI implementation may have a negative impact on universities that will not see any payoff from such technologies (Bucea-Manea-Toniș et al., 2022).

Digital Inequity

Digital inequity is another relevant issue when discussing AI implementation in Romania. AI in Romanian higher education should not only contribute to cost reduction for students but also provide opportunities for open access to higher education and foster digital inclusion, ensuring that individuals are not left behind in the offline world (Grosseck, 2019).


Bucea-Manea-Țoniş, Rocsana, Valentin Kuleto, Simona Corina Dobre Gudei, Costin Lianu, Cosmin Lianu, Milena P. Ilić, and Dan Păun. „Artificial Intelligence Potential in Higher Education Institutions Enhanced Learning Environment in Romania and Serbia.” Sustainability 14, no. 10 (2022): 5842.

Chan, Cecilia Ka Yuk, and Wenjie Hu. „Students’ Voices on Generative AI: Perceptions, Benefits, and Challenges in Higher Education.” arXiv:2305.00290, (2023).

Dave, Manas, and Neil Patel. „Artificial Intelligence in Healthcare and Education.” British Dental Journal 234, no. 10 (2023): 761-764.

Grosseck, Gabriela, Laura Malița, and Ramona Bran. „Digital University-Issues and Trends in Romanian Higher Education.” Broad Research in Artificial Intelligence and Neuroscience 10, no. 1 (2019): 108-122.

Owoc, Mieczysław L., Agnieszka Sawicka, and Paweł Weichbroth. „Artificial Intelligence Technologies in Education: Benefits, Challenges and Strategies of Implementation.” In Artificial Intelligence for Knowledge Management: 7th IFIP WG 12.6 International Workshop, AI4KM 2019, Held at IJCAI 2019, Macao, China, August 11, 2019, Revised Selected Papers, pp. 37-58. Cham: Springer International Publishing, 2021.

Pisica, Alina Iorga, Tudor Edu, Rodica Milena Zaharia, and Razvan Zaharia. „Implementing Artificial Intelligence in Higher Education: Pros and Cons from the Perspectives of Academics.” Societies 13, no. 5 (2023): 118.

Rasul, Tareq, Sumesh Nair, Diane Kalendra, Mulyadi Robin, Fernando de Oliveira Santini, Wagner Junior Ladeira, Mingwei Sun, Ingrid Day, Raouf Ahmad Rather, and Liz Heathcote. „The Role of ChatGPT in Higher Education: Benefits, Challenges, and Future Research Directions.” Journal of Applied Learning and Teaching 6, no. 1 (2023).

Sallam, Malik, Nesreen Salim, Muna Barakat, and Alaa Al-Tammemi. „ChatGPT Applications in Medical, Dental, Pharmacy, and Public Health Education: A Descriptive Study Highlighting the Advantages and Limitations.” Narra J 3, no. 1 (2023).