24°C
Study Finds Readers Prefer AI-Generated Stories When They Believe Humans Wrote Them
New research from Villanova University suggests readers consistently give higher ratings to stories generated by artificial intelligence than to human-written fiction—particularly when they mistakenly believe the author is a person.
Readers tend to rate AI-generated stories more favorably than fiction written by human authors when they believe the work was created by a person, according to a new study by researchers at Villanova University. The findings suggest that perceptions about authorship can significantly influence how people judge the quality and appeal of written content.
Published in the journal Judgment and Decision Making, the research also found that most participants struggled to distinguish between stories written by humans and those produced by artificial intelligence. The results add to the growing debate over how AI-generated content is perceived as generative AI tools become increasingly common across publishing, education, and digital media.
Three Experiments Tested Reader Perceptions
The researchers recruited adult participants between the ages of 18 and 81 to evaluate whether they could identify the origin of short fictional stories, compare their quality, and determine how engaging they found them. The study used three short stories written by human authors that had appeared in respected literary publications alongside three stories generated by ChatGPT.
In the first experiment, 1,682 participants read a single story and were told—accurately or inaccurately—that it had been written either by a human or by an AI system. They then rated its quality and overall appeal. In the second and third experiments, participants read one human-written story and one AI-generated story without being told their origins before attempting to identify the author of each.
Clearer Writing Received Higher Ratings
Across the experiments, AI-generated stories consistently received higher scores than human-written fiction, with the strongest preference appearing when readers mistakenly believed the AI-created text had been written by a person. According to the researchers, this pattern points to a favorable bias toward human authorship even when readers ultimately preferred the writing style produced by artificial intelligence.
The study suggests that AI-generated narratives may appeal to readers because they are typically more direct, easier to follow, and more predictable than stories written by human authors, which often rely on greater subtlety and complexity.
AI Familiarity Improved Detection
The research also examined whether participants’ backgrounds influenced their ability to recognize AI-generated writing. Those who reported greater familiarity with artificial intelligence were more successful at identifying texts created by AI systems.
By contrast, participants with more experience reading fiction were no better at distinguishing between human and AI authors. The findings indicate that technical familiarity with AI may be a stronger predictor of detection accuracy than literary expertise.
Deena Weisberg, the study’s lead author and a researcher in the Department of Psychological and Brain Sciences at Villanova University, said AI-generated writing generally tends to be clearer, more direct, and easier to process, while human-authored fiction often incorporates greater nuance and complexity that may not always translate into higher reader ratings.
