Understanding the impact of AI Hallucinations on the university community

Abstract

Since we live in the era of the information revolution, finding trusted and accurate information takes time and effort made students and researchers aim to find an easier way. Generative AI (Artificial Intelligence) tools offer an easy solution for accessing the required information easily and accessible; however, these tools rely on vast datasets to predict statistically probable outputs, not guaranteed ac- curacy. This can lead to misinformation, factual errors, biases, and fabricated content, which is termed "hallucinations." The research problem focuses on the challenges of detecting these AI hallucinations, the main issue for all users of AI technologies. The main objective of the study is to raise awareness about AI hallucinations and promote the ethical and effective use of AI tools among New Giza University students, faculty, and staff. This involves the approach to understanding the biases and errors associated with AI outputs. Methodologically, the study will employ a mixed-methods approach, combining quantitative analyses of AI tool accuracy with collecting qualitative data via survey of users across a range of fields to gather insights on the impact of AI hallucinations. The expected results of this research are to reveal the pitfalls that researchers might run into when relying on AI technology for their work. Additionally, the findings will contribute significantly to information literacy programs, by advocating for the including of AI tool assessments within the broader information literacy curriculum and equipping users with the skills to critically evaluate AI-generated content.