Published December 2025
| Version v1
Thesis
Algorithmic Persuasion: Exploring the Role of Sentiment and "Quotation" in AI‐Generated Review Summaries
Description
The rapid rise of AI-generated summaries (AIGS) is transforming how consumers navigate online review environments. While prior work has highlighted platform-level outcomes, little is known about its persuasive effects. To address this gap, we investigate how sentiment orientation and the presence of quotation marks in AI-generated review summaries influence consumer evaluation and behavioral intentions. Across four preregistered experiments (N > 3,000), we find that AI-generated summaries that emphasize positive sentiment inflate predicted ratings and narrow perceived differences between high- and low-rated restaurants, while quotation marks enhance evaluations but reduce review exploration, particularly when other evaluative cues (e.g., star ratings or access to individual reviews) are unavailable. Together, these findings demonstrate that the linguistic and stylistic cues in AIGS can meaningfully influence consumer judgment and decision-making.
Additional details
Identifiers
- Other
- oai:uchicago.tind.io:16684