Loading...
Paper Type
Complete
Abstract
Generative Artificial Intelligence has become one of the most discussed topics in consumer-oriented Artificial Intelligence since 2022, often being compared to human intelligence in various domains. In this study, we examine the implications of these technologies for the field of fake review detection through an experimental approach that encompasses both the generation and detection of fake reviews. Our goal is to provide a well-founded assessment of the current technological threat of AI-generated fake reviews. We employ state-of-the-art models for review generation and evaluate the generated content using both open-source and subscription-based models. Our findings suggest that the threat posed by AI-generated fake reviews is currently considered low, especially concerning the quality of textual reviews. Interestingly, our analysis also raises the question of whether an adaptation of already established models is indeed necessary by incorporating generated fake reviews by Large Language Models in the model training process.
Paper Number
1132
Recommended Citation
Theuerkauf, René and Peters, Ralf, "Battlefield of Online Product Reviews: AI vs. AI" (2024). AMCIS 2024 Proceedings. 15.
https://aisel.aisnet.org/amcis2024/adoptdiff/adoptdiff/15
Battlefield of Online Product Reviews: AI vs. AI
Generative Artificial Intelligence has become one of the most discussed topics in consumer-oriented Artificial Intelligence since 2022, often being compared to human intelligence in various domains. In this study, we examine the implications of these technologies for the field of fake review detection through an experimental approach that encompasses both the generation and detection of fake reviews. Our goal is to provide a well-founded assessment of the current technological threat of AI-generated fake reviews. We employ state-of-the-art models for review generation and evaluate the generated content using both open-source and subscription-based models. Our findings suggest that the threat posed by AI-generated fake reviews is currently considered low, especially concerning the quality of textual reviews. Interestingly, our analysis also raises the question of whether an adaptation of already established models is indeed necessary by incorporating generated fake reviews by Large Language Models in the model training process.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
SIGADIT