The scientific paper entitled “Exploring the Challenges and Potential of Generative AI: Insights from an Empirical Study” has been published in open access. Carried out within the OpenFact project, the research highlights both the capabilities and the practical challenges posed by these advanced AI tools.
The experiment results revealed frequent occurrences of AI “hallucinations”, instances where AI generates convincingly realistic yet incorrect or nonexistent information. Participants also noted a major issue concerning the transparency of AI-generated content, specifically the lack of citations or identifiable sources, making it difficult to verify the credibility of the information provided.
Another alarming finding was how easily generative AI tools could be manipulated to produce fake news. Students participating in the experiment found it remarkably simple to prompt AI into generating convincing false information, which could have significant social and economic repercussions.
These findings have profound implications for businesses, educational institutions, and the broader sphere of information integrity. Companies integrating AI technologies must carefully assess the risks associated with unreliable AI-generated information, which could lead to poor business decisions or damage to their reputation. Similarly, academic institutions must address the impacts on academic integrity and the quality of student-produced content when AI tools are utilized.
The authors emphasize that the rapid advancement of generative AI technology makes conducting this type of research challenging yet crucial. Continued study and vigilance are necessary to ensure that the development of AI tools remains responsible, transparent, and beneficial to society.
Authors of the paper: Dr. Milena Stróżyna, Prof. Krzysztof Węcel, Dr. Piotr Stolarski, Ewelina Księżniak, Marcin Sawiński, Dr. Włodzimierz Lewoniewski, Prof. Witold Abramowicz.
More information can be found on the websites: kie.ue.poznan.pl, ue.poznan.pl