Global University Student Survey Uncovers ChatGPT’s Strengths and Shortcomings

A new worldwide study of over 23,000 university students explores ChatGPT’s benefits, limitations, and major ethical implications for higher education.

• Large-scale survey encompassed 23,218 students across 109 countries
• Many recognise ChatGPT’s academic value, but question reliability and ethics
• Findings underscore varying impacts across diverse socioeconomic contexts

In new research published in PLOS One an international team led by Dejan Ravšelj of the University of Ljubljana surveyed 23,218 higher education students from 109 countries to gauge their perceptions of ChatGPT. Gathering data through an anonymous online questionnaire, the researchers explored usage habits, ethical concerns, and the chatbot’s potential impact on employability. Conducted between October 2023 and February 2024, the survey aimed to capture a global snapshot of how students use and experience the AI chatbot. Participants largely view ChatGPT as a helpful tool, especially for summarising complex information and drafting academic texts, yet the findings reveal persistent concerns around reliability, ethics, and its influence on critical thinking.

Positive Views Balanced by Ethical Fears

Overall, a majority of students expressed favourable attitudes, with 70% describing ChatGPT as interesting to use. However, only 29% reported using it for brainstorming, and a mere 11% turned to ChatGPT for creative writing. Many students see it as an efficient means to simplify complicated concepts and assist with academic composition. At the same time, reliability issues and possible erosion of critical thinking were frequently cited as reasons for caution. The authors state: “A new milestone in generative artificial intelligence research in higher education: Early global perceptions of ChatGPT, drawn from over 23,000 students across 109 countries and territories, highlight its role as an engaging learning tool while raising ethical concerns that require further attention. The study examines student perspectives on its usage, capabilities, regulation, satisfaction, attitudes, study outcomes, skills development, labour market implications, and emotional responses.”

In line with these findings, concerns about academic misconduct and privacy have emerged, suggesting that institutional guidelines may need to adapt to ChatGPT’s growing role. Many respondents cited worries over the chatbot’s tendency to generate flawed or misleading responses, reinforcing the importance of fact-checking when relying on AI-driven resources. Although 25% of students found it easier to interact with ChatGPT than with peers, a preference for human engagement prevails among the majority. When asked about ideal usage scenarios, many suggested adopting ChatGPT as a supplemental learning aid rather than a replacement for traditional pedagogy. Nonetheless, there is mounting awareness that AI familiarity could become an increasingly essential skill in the future labour market, particularly in academia.

Socioeconomic Influences and Future Directions

Interestingly, the survey suggests distinct usage patterns based on economic context. Students in lower-income regions emphasised ChatGPT’s ability to fill gaps in educational resources, while peers in higher-income countries focused more on its advanced capabilities. This divergence points to a need for inclusive policy development, ensuring that AI technologies bolster learning without exacerbating existing inequalities. The researchers highlight that ongoing investigations should track students’ perceptions over time and expand participation in underrepresented regions. By recognising these nuances, higher education institutions can tailor AI adoption strategies that both harness benefits and address legitimate concerns.

As the results illuminate both advantages and drawbacks, the authors underscore the need for balanced, evidence-based approaches to AI usage in the classroom. They point out that these findings “could help inform the design of higher education curricula and policies to harness the benefits of ChatGPT equitably across diverse student populations.” Such measures might involve training educators to integrate chatbots responsibly, establishing transparency around data privacy, and clearly defining acceptable usage guidelines. With these insights, academic leaders and policymakers are better equipped to develop frameworks that maximise ChatGPT’s educational potential, while mitigating ethical risks and preserving the core values of rigorous scholarly training.