Design AI-Generated Summaries for Online Video Platforms: Evidence from a Field Experiment

Abstract

AI-generated summaries condense lengthy videos into concise text, helping reduce information overload and guide user attention. This research designs two summarization strategies for online videos: information-extractive AI-generated summaries (IAIGS), which present fact-based recaps, and suspense-inducing AI-generated summaries (SAIGS), which withhold key details to spark curiosity. To assess their impact on video consumption, we conducted a randomized field experiment on a video-sharing platform, focusing on two content genres: Science (instrumental needs), and Humanities & History (affective needs). Our results show that IAIGS reduced video views across both genres, driven by information substitution. In contrast, SAIGS have genre-specific effects: they increase engagement with Humanities & History content by sparking curiosity but decrease engagement with Science content by obstructing information-seeking. Overall, our study underscores that summary design shapes content consumption and that the effectiveness of AI-generated summaries depends on the dynamics of human–AI interaction. The findings provide practical insights for designing summaries to enhance user engagement and content consumption.