Authors: B Lebrun, S Temtsin, A Vonasch
Year: 2024
Published in: Frontiers in Robotics and ..., 2024 - frontiersin.org
Institution: University of Lausanne, University of California Berkeley, University of Massachusetts Amherst, Arizona State University
Research Area: AI in Social Science Research, Survey Methodology, Data Quality
Discipline: Artificial Intelligence
The study examines the integrity of online questionnaire responses and concludes that humans can identify AI-generated text with 76% accuracy, but current AI detection systems are ineffective, raising concerns about data quality in online surveys.
Methods: Human participants and automatic AI detection systems were tested on their ability to differentiate AI-generated text from human-generated text in the context of online questionnaires.
Key Findings: The study measured the ability of humans and AI detection tools to correctly identify whether text was generated by a human or an AI system for online questionnaire responses.
DOI: https://doi.org/10.3389/frobt.2023.1277635
Citations: 26